diff --git a/README.md b/README.md new file mode 100644 index 0000000..4417fc3 --- /dev/null +++ b/README.md @@ -0,0 +1,18 @@ +# Machine Learning Project – Summer Semester 2025 + +This project was created as part of the "Machine Learning" course at HTW Saar in the Practical Computer Science study program. + +## Objective + +We are developing a Jupyter Notebook that automatically predicts the genre of Steam games based on their descriptions. +As a data basis, we use a publicly available Steam Games dataset that we found on Kaggle. + +## Dataset + +We use the [Steam Games Dataset from Kaggle](https://www.kaggle.com/datasets/artermiloff/steam-games-dataset/data). + +## Contributors + +- Maximilian Kany +- Florian Speicher +- Tim Wall \ No newline at end of file diff --git a/comparison.py b/comparison.py new file mode 100644 index 0000000..fcced39 --- /dev/null +++ b/comparison.py @@ -0,0 +1,140 @@ +import os +import numpy as np +import pandas as pd +from sklearn import set_config + +from sklearn.compose import ColumnTransformer +from sklearn.preprocessing import FunctionTransformer + +from sklearn.preprocessing import MultiLabelBinarizer +import ast + + +from sklearn.feature_extraction.text import TfidfVectorizer +from sklearn.linear_model import LogisticRegression +from sklearn.multioutput import MultiOutputClassifier +from sklearn.metrics import classification_report +from sklearn.model_selection import train_test_split +from sklearn.datasets import load_iris +from sklearn.metrics import accuracy_score, classification_report +from sklearn.svm import SVC, LinearSVC +from sklearn.tree import DecisionTreeClassifier +from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier +from sklearn.naive_bayes import GaussianNB, MultinomialNB, BernoulliNB +from sklearn.neighbors import KNeighborsClassifier +from sklearn.neural_network import MLPClassifier + + +set_config(transform_output="pandas") # dataframe supremacy + +def prepDataset(dataset): #returns X_train, X_test, y_train, y_test + dataset = pd.read_csv("./games_march2025_cleaned_2k.csv",sep=",") + # desc, genres, tags + column_transformer = ColumnTransformer([ + # merge all descriptions + ('desc', FunctionTransformer(lambda X: X.fillna('').agg(' '.join, axis=1).to_frame(name="desc")), + ['detailed_description', 'about_the_game', 'short_description']), + ('pass', 'passthrough', ['genres']),#, 'tags' + ], + verbose_feature_names_out=False + ) + dataset = column_transformer.fit_transform(dataset) + + + + #### SET MISSING VALUES + print("SETMISS") + # Setting missing numeric values to the mean + dataset.fillna(dataset.mean(numeric_only=True), inplace=True) + # Setting missing text values to 'Unknown' + dataset.fillna('', inplace=True) + # Setting missing values in other columns to NaN + dataset.dropna(inplace=True) + + ##### STRUCTURIZE GENRES to onehot + #serialize array + dataset['genres'] = dataset['genres'].map(lambda s: ast.literal_eval(s)) + #print(dataset['genres']) # in py but not yet onehotenc + + # MultiLabelBinarizer does onehotenc for arrays + mlb_genres = MultiLabelBinarizer() + genres_encoded = mlb_genres.fit_transform(dataset.pop('genres')) + #genres_count = len(mlb_genres.classes_) # for multi-label classifiction later + + genres_df = pd.DataFrame(genres_encoded, columns=mlb_genres.classes_) + #print(genres_df) + #dataset = pd.concat([dataset, genres_df], axis=1) + #print(dataset) + + + #### convert text to bag of words + + ## Count vs Tfidf vectorizer + vectorizer = TfidfVectorizer() + tfidf_matrix = vectorizer.fit_transform(dataset['desc']) # matrix + tfidf_df = pd.DataFrame(tfidf_matrix.toarray(), columns=vectorizer.get_feature_names_out()) + #print(tfidf_df) + + + ##### MODEL + print("MODEL") + + + X = tfidf_df + y = genres_df + # cleanup datapoints that dont have a target value (all target columns are 0) + mask = y.sum(axis=1).map(lambda x: x > 0) + #print((mask == False).sum()) #31 cases with all target columns 0 + X_clean = X[mask] + y_clean = y[mask] + + # Split dataset + return train_test_split(X_clean, y_clean, random_state=0) + +def comparison(X_train, X_test, y_train, y_test, estimator, jobs: int = 1): #returns class_report + multi_target_clf = MultiOutputClassifier(estimator, n_jobs=jobs) # LogisticRegression(max_iter=1337, random_state=0) + + # model training + multi_target_clf.fit(X_train, y_train) + + # predict against test data + y_pred = multi_target_clf.predict(X_test) + return classification_report(y_test, y_pred, zero_division=0.0) + +datasets = [ + 'games_march2025_cleaned_2k.csv', + 'games_march2025_cleaned_10k.csv', + 'games_march2025_cleaned.csv' +] + +estimators = { + "LogisticRegression-i1000": LogisticRegression(max_iter=1000, random_state=0), + "LogisticRegression-i10000": LogisticRegression(max_iter=10000, random_state=0), + "LinearSVC-i5000": LinearSVC(max_iter=5000), + "SVC-RBF-i10000": SVC(kernel="rbf", max_iter=10000), + "DecisionTreeClassifier": DecisionTreeClassifier(random_state=0), + "RandomForestClassifier": RandomForestClassifier(random_state=0), + "GradientBoostingClassifier": GradientBoostingClassifier(random_state=0), + "GaussianNB": GaussianNB(), + "MultinomialNB": MultinomialNB(), + "BernoulliNB": BernoulliNB(), + "MLPClassifier-i10000": MLPClassifier(max_iter=10000, random_state=0), +} + +for dataset in datasets: + print("-" * 60) + print("dataset -> " + dataset) + print("-" * 60) + print("mkdir") + folder = dataset.split(".csv")[0] + if not os.path.isdir(folder): + os.mkdir(folder) + X_train, X_test, y_train, y_test = prepDataset(dataset) + for esti in estimators: + compari = comparison(X_train, X_test, y_train, y_test, estimators[esti], 1) #TODO: change the job count if you can + print("open") + f = open(folder + "/" + esti +".txt", mode="w+", encoding="utf-8") + f.write(compari) + print("write") + f.close() + print("close") \ No newline at end of file diff --git a/games_march2025_cleaned/BernoulliNB.txt b/games_march2025_cleaned/BernoulliNB.txt new file mode 100644 index 0000000..f2237d4 --- /dev/null +++ b/games_march2025_cleaned/BernoulliNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.75 0.90 0.82 300 + 1 0.72 0.68 0.70 216 + 2 0.50 0.08 0.14 86 + 3 0.27 0.07 0.11 46 + 4 0.40 0.07 0.12 83 + 5 0.00 0.00 0.00 0 + 6 0.77 0.82 0.79 245 + 7 0.33 0.10 0.15 42 + 8 0.67 0.40 0.50 127 + 9 0.00 0.00 0.00 12 + 10 0.71 0.37 0.49 127 + 11 0.00 0.00 0.00 14 + 12 0.49 0.31 0.38 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.70 0.55 0.62 1404 + macro avg 0.40 0.27 0.30 1404 +weighted avg 0.64 0.55 0.56 1404 + samples avg 0.73 0.59 0.61 1404 diff --git a/games_march2025_cleaned/DecisionTreeClassifier.txt b/games_march2025_cleaned/DecisionTreeClassifier.txt new file mode 100644 index 0000000..900c256 --- /dev/null +++ b/games_march2025_cleaned/DecisionTreeClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.76 0.73 0.75 300 + 1 0.56 0.53 0.54 216 + 2 0.36 0.33 0.34 86 + 3 0.33 0.26 0.29 46 + 4 0.40 0.46 0.43 83 + 5 0.00 0.00 0.00 0 + 6 0.65 0.61 0.63 245 + 7 0.39 0.40 0.40 42 + 8 0.59 0.57 0.58 127 + 9 0.60 0.25 0.35 12 + 10 0.56 0.51 0.53 127 + 11 0.39 0.50 0.44 14 + 12 0.52 0.49 0.50 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.58 0.55 0.57 1404 + macro avg 0.44 0.40 0.41 1404 +weighted avg 0.58 0.55 0.57 1404 + samples avg 0.59 0.59 0.55 1404 diff --git a/games_march2025_cleaned/GaussianNB.txt b/games_march2025_cleaned/GaussianNB.txt new file mode 100644 index 0000000..83d7a2e --- /dev/null +++ b/games_march2025_cleaned/GaussianNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.76 0.80 0.78 300 + 1 0.62 0.51 0.56 216 + 2 0.63 0.14 0.23 86 + 3 0.17 0.02 0.04 46 + 4 0.42 0.10 0.16 83 + 5 0.00 0.00 0.00 0 + 6 0.68 0.66 0.67 245 + 7 0.56 0.12 0.20 42 + 8 0.55 0.33 0.41 127 + 9 0.67 0.17 0.27 12 + 10 0.65 0.31 0.42 127 + 11 1.00 0.14 0.25 14 + 12 0.53 0.29 0.38 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.66 0.47 0.55 1404 + macro avg 0.52 0.26 0.31 1404 +weighted avg 0.62 0.47 0.51 1404 + samples avg 0.67 0.53 0.55 1404 diff --git a/games_march2025_cleaned/GradientBoostingClassifier.txt b/games_march2025_cleaned/GradientBoostingClassifier.txt new file mode 100644 index 0000000..7c8ce6e --- /dev/null +++ b/games_march2025_cleaned/GradientBoostingClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.85 0.80 0.83 300 + 1 0.77 0.61 0.68 216 + 2 0.55 0.13 0.21 86 + 3 0.42 0.11 0.17 46 + 4 0.68 0.33 0.44 83 + 5 0.00 0.00 0.00 0 + 6 0.71 0.76 0.74 245 + 7 0.61 0.26 0.37 42 + 8 0.81 0.50 0.61 127 + 9 0.75 0.25 0.38 12 + 10 0.81 0.54 0.65 127 + 11 0.40 0.43 0.41 14 + 12 0.69 0.42 0.53 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.76 0.57 0.65 1404 + macro avg 0.57 0.37 0.43 1404 +weighted avg 0.74 0.57 0.63 1404 + samples avg 0.76 0.63 0.65 1404 diff --git a/games_march2025_cleaned/LinearSVC-i5000.txt b/games_march2025_cleaned/LinearSVC-i5000.txt new file mode 100644 index 0000000..df82b40 --- /dev/null +++ b/games_march2025_cleaned/LinearSVC-i5000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.85 0.87 0.86 300 + 1 0.76 0.66 0.70 216 + 2 0.77 0.20 0.31 86 + 3 0.00 0.00 0.00 46 + 4 0.76 0.27 0.39 83 + 5 0.00 0.00 0.00 0 + 6 0.78 0.81 0.79 245 + 7 0.89 0.19 0.31 42 + 8 0.77 0.60 0.67 127 + 9 1.00 0.58 0.74 12 + 10 0.85 0.54 0.66 127 + 11 1.00 0.29 0.44 14 + 12 0.82 0.42 0.56 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.80 0.61 0.69 1404 + macro avg 0.66 0.39 0.46 1404 +weighted avg 0.78 0.61 0.66 1404 + samples avg 0.81 0.67 0.69 1404 diff --git a/games_march2025_cleaned/LogisticRegression-i1000.txt b/games_march2025_cleaned/LogisticRegression-i1000.txt new file mode 100644 index 0000000..b7926d4 --- /dev/null +++ b/games_march2025_cleaned/LogisticRegression-i1000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.78 0.91 0.84 300 + 1 0.78 0.62 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.04 0.07 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.81 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.90 0.34 0.49 127 + 9 0.00 0.00 0.00 12 + 10 0.89 0.25 0.39 127 + 11 0.00 0.00 0.00 14 + 12 0.88 0.14 0.24 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.50 0.61 1404 + macro avg 0.50 0.22 0.26 1404 +weighted avg 0.77 0.50 0.53 1404 + samples avg 0.77 0.56 0.60 1404 diff --git a/games_march2025_cleaned/LogisticRegression-i10000.txt b/games_march2025_cleaned/LogisticRegression-i10000.txt new file mode 100644 index 0000000..b7926d4 --- /dev/null +++ b/games_march2025_cleaned/LogisticRegression-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.78 0.91 0.84 300 + 1 0.78 0.62 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.04 0.07 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.81 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.90 0.34 0.49 127 + 9 0.00 0.00 0.00 12 + 10 0.89 0.25 0.39 127 + 11 0.00 0.00 0.00 14 + 12 0.88 0.14 0.24 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.50 0.61 1404 + macro avg 0.50 0.22 0.26 1404 +weighted avg 0.77 0.50 0.53 1404 + samples avg 0.77 0.56 0.60 1404 diff --git a/games_march2025_cleaned/MLPClassifier-i10000.txt b/games_march2025_cleaned/MLPClassifier-i10000.txt new file mode 100644 index 0000000..c4634dc --- /dev/null +++ b/games_march2025_cleaned/MLPClassifier-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.84 0.85 0.84 300 + 1 0.73 0.67 0.70 216 + 2 0.74 0.30 0.43 86 + 3 0.50 0.02 0.04 46 + 4 0.69 0.24 0.36 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.79 0.79 245 + 7 0.86 0.14 0.24 42 + 8 0.76 0.63 0.69 127 + 9 1.00 0.33 0.50 12 + 10 0.81 0.52 0.63 127 + 11 1.00 0.14 0.25 14 + 12 0.75 0.41 0.53 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.60 0.68 1404 + macro avg 0.68 0.36 0.43 1404 +weighted avg 0.78 0.60 0.65 1404 + samples avg 0.80 0.66 0.68 1404 diff --git a/games_march2025_cleaned/MultinomialNB.txt b/games_march2025_cleaned/MultinomialNB.txt new file mode 100644 index 0000000..bc74cf3 --- /dev/null +++ b/games_march2025_cleaned/MultinomialNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.64 0.99 0.78 300 + 1 0.85 0.24 0.37 216 + 2 0.60 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 0.80 0.05 0.09 83 + 5 0.00 0.00 0.00 0 + 6 0.78 0.80 0.79 245 + 7 0.40 0.05 0.09 42 + 8 1.00 0.04 0.08 127 + 9 0.00 0.00 0.00 12 + 10 0.20 0.01 0.02 127 + 11 0.00 0.00 0.00 14 + 12 1.00 0.05 0.09 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.69 0.40 0.51 1404 + macro avg 0.45 0.16 0.17 1404 +weighted avg 0.68 0.40 0.39 1404 + samples avg 0.70 0.44 0.50 1404 diff --git a/games_march2025_cleaned/RandomForestClassifier.txt b/games_march2025_cleaned/RandomForestClassifier.txt new file mode 100644 index 0000000..6fbe546 --- /dev/null +++ b/games_march2025_cleaned/RandomForestClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.80 0.88 0.84 300 + 1 0.78 0.55 0.64 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.06 0.11 83 + 5 0.00 0.00 0.00 0 + 6 0.74 0.78 0.76 245 + 7 0.00 0.00 0.00 42 + 8 0.84 0.24 0.38 127 + 9 0.00 0.00 0.00 12 + 10 0.91 0.24 0.38 127 + 11 1.00 0.14 0.25 14 + 12 1.00 0.25 0.39 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.48 0.59 1404 + macro avg 0.58 0.23 0.27 1404 +weighted avg 0.78 0.48 0.52 1404 + samples avg 0.77 0.54 0.60 1404 diff --git a/games_march2025_cleaned/SVC-RBF-i10000.txt b/games_march2025_cleaned/SVC-RBF-i10000.txt new file mode 100644 index 0000000..ff0c7b7 --- /dev/null +++ b/games_march2025_cleaned/SVC-RBF-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.81 0.90 0.85 300 + 1 0.76 0.63 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.05 0.09 83 + 5 0.00 0.00 0.00 0 + 6 0.77 0.83 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.84 0.40 0.54 127 + 9 1.00 0.17 0.29 12 + 10 0.90 0.34 0.49 127 + 11 1.00 0.14 0.25 14 + 12 0.92 0.21 0.34 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.80 0.53 0.63 1404 + macro avg 0.64 0.26 0.32 1404 +weighted avg 0.79 0.53 0.56 1404 + samples avg 0.79 0.59 0.63 1404 diff --git a/games_march2025_cleaned_10k.csv b/games_march2025_cleaned_10k.csv new file mode 100644 index 0000000..2c3c073 --- /dev/null +++ b/games_march2025_cleaned_10k.csv @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:12cf598a6e41d83cfa9c16e99d4d9578cb4ee7c3594fae9f9b921772887a08d7 +size 68658136 diff --git a/games_march2025_cleaned_10k/BernoulliNB.txt b/games_march2025_cleaned_10k/BernoulliNB.txt new file mode 100644 index 0000000..f2237d4 --- /dev/null +++ b/games_march2025_cleaned_10k/BernoulliNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.75 0.90 0.82 300 + 1 0.72 0.68 0.70 216 + 2 0.50 0.08 0.14 86 + 3 0.27 0.07 0.11 46 + 4 0.40 0.07 0.12 83 + 5 0.00 0.00 0.00 0 + 6 0.77 0.82 0.79 245 + 7 0.33 0.10 0.15 42 + 8 0.67 0.40 0.50 127 + 9 0.00 0.00 0.00 12 + 10 0.71 0.37 0.49 127 + 11 0.00 0.00 0.00 14 + 12 0.49 0.31 0.38 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.70 0.55 0.62 1404 + macro avg 0.40 0.27 0.30 1404 +weighted avg 0.64 0.55 0.56 1404 + samples avg 0.73 0.59 0.61 1404 diff --git a/games_march2025_cleaned_10k/DecisionTreeClassifier.txt b/games_march2025_cleaned_10k/DecisionTreeClassifier.txt new file mode 100644 index 0000000..900c256 --- /dev/null +++ b/games_march2025_cleaned_10k/DecisionTreeClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.76 0.73 0.75 300 + 1 0.56 0.53 0.54 216 + 2 0.36 0.33 0.34 86 + 3 0.33 0.26 0.29 46 + 4 0.40 0.46 0.43 83 + 5 0.00 0.00 0.00 0 + 6 0.65 0.61 0.63 245 + 7 0.39 0.40 0.40 42 + 8 0.59 0.57 0.58 127 + 9 0.60 0.25 0.35 12 + 10 0.56 0.51 0.53 127 + 11 0.39 0.50 0.44 14 + 12 0.52 0.49 0.50 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.58 0.55 0.57 1404 + macro avg 0.44 0.40 0.41 1404 +weighted avg 0.58 0.55 0.57 1404 + samples avg 0.59 0.59 0.55 1404 diff --git a/games_march2025_cleaned_10k/GaussianNB.txt b/games_march2025_cleaned_10k/GaussianNB.txt new file mode 100644 index 0000000..83d7a2e --- /dev/null +++ b/games_march2025_cleaned_10k/GaussianNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.76 0.80 0.78 300 + 1 0.62 0.51 0.56 216 + 2 0.63 0.14 0.23 86 + 3 0.17 0.02 0.04 46 + 4 0.42 0.10 0.16 83 + 5 0.00 0.00 0.00 0 + 6 0.68 0.66 0.67 245 + 7 0.56 0.12 0.20 42 + 8 0.55 0.33 0.41 127 + 9 0.67 0.17 0.27 12 + 10 0.65 0.31 0.42 127 + 11 1.00 0.14 0.25 14 + 12 0.53 0.29 0.38 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.66 0.47 0.55 1404 + macro avg 0.52 0.26 0.31 1404 +weighted avg 0.62 0.47 0.51 1404 + samples avg 0.67 0.53 0.55 1404 diff --git a/games_march2025_cleaned_10k/GradientBoostingClassifier.txt b/games_march2025_cleaned_10k/GradientBoostingClassifier.txt new file mode 100644 index 0000000..7c8ce6e --- /dev/null +++ b/games_march2025_cleaned_10k/GradientBoostingClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.85 0.80 0.83 300 + 1 0.77 0.61 0.68 216 + 2 0.55 0.13 0.21 86 + 3 0.42 0.11 0.17 46 + 4 0.68 0.33 0.44 83 + 5 0.00 0.00 0.00 0 + 6 0.71 0.76 0.74 245 + 7 0.61 0.26 0.37 42 + 8 0.81 0.50 0.61 127 + 9 0.75 0.25 0.38 12 + 10 0.81 0.54 0.65 127 + 11 0.40 0.43 0.41 14 + 12 0.69 0.42 0.53 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.76 0.57 0.65 1404 + macro avg 0.57 0.37 0.43 1404 +weighted avg 0.74 0.57 0.63 1404 + samples avg 0.76 0.63 0.65 1404 diff --git a/games_march2025_cleaned_10k/LinearSVC-i5000.txt b/games_march2025_cleaned_10k/LinearSVC-i5000.txt new file mode 100644 index 0000000..df82b40 --- /dev/null +++ b/games_march2025_cleaned_10k/LinearSVC-i5000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.85 0.87 0.86 300 + 1 0.76 0.66 0.70 216 + 2 0.77 0.20 0.31 86 + 3 0.00 0.00 0.00 46 + 4 0.76 0.27 0.39 83 + 5 0.00 0.00 0.00 0 + 6 0.78 0.81 0.79 245 + 7 0.89 0.19 0.31 42 + 8 0.77 0.60 0.67 127 + 9 1.00 0.58 0.74 12 + 10 0.85 0.54 0.66 127 + 11 1.00 0.29 0.44 14 + 12 0.82 0.42 0.56 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.80 0.61 0.69 1404 + macro avg 0.66 0.39 0.46 1404 +weighted avg 0.78 0.61 0.66 1404 + samples avg 0.81 0.67 0.69 1404 diff --git a/games_march2025_cleaned_10k/LogisticRegression-i1000.txt b/games_march2025_cleaned_10k/LogisticRegression-i1000.txt new file mode 100644 index 0000000..b7926d4 --- /dev/null +++ b/games_march2025_cleaned_10k/LogisticRegression-i1000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.78 0.91 0.84 300 + 1 0.78 0.62 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.04 0.07 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.81 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.90 0.34 0.49 127 + 9 0.00 0.00 0.00 12 + 10 0.89 0.25 0.39 127 + 11 0.00 0.00 0.00 14 + 12 0.88 0.14 0.24 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.50 0.61 1404 + macro avg 0.50 0.22 0.26 1404 +weighted avg 0.77 0.50 0.53 1404 + samples avg 0.77 0.56 0.60 1404 diff --git a/games_march2025_cleaned_10k/LogisticRegression-i10000.txt b/games_march2025_cleaned_10k/LogisticRegression-i10000.txt new file mode 100644 index 0000000..b7926d4 --- /dev/null +++ b/games_march2025_cleaned_10k/LogisticRegression-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.78 0.91 0.84 300 + 1 0.78 0.62 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.04 0.07 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.81 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.90 0.34 0.49 127 + 9 0.00 0.00 0.00 12 + 10 0.89 0.25 0.39 127 + 11 0.00 0.00 0.00 14 + 12 0.88 0.14 0.24 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.50 0.61 1404 + macro avg 0.50 0.22 0.26 1404 +weighted avg 0.77 0.50 0.53 1404 + samples avg 0.77 0.56 0.60 1404 diff --git a/games_march2025_cleaned_10k/MLPClassifier-i10000.txt b/games_march2025_cleaned_10k/MLPClassifier-i10000.txt new file mode 100644 index 0000000..c4634dc --- /dev/null +++ b/games_march2025_cleaned_10k/MLPClassifier-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.84 0.85 0.84 300 + 1 0.73 0.67 0.70 216 + 2 0.74 0.30 0.43 86 + 3 0.50 0.02 0.04 46 + 4 0.69 0.24 0.36 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.79 0.79 245 + 7 0.86 0.14 0.24 42 + 8 0.76 0.63 0.69 127 + 9 1.00 0.33 0.50 12 + 10 0.81 0.52 0.63 127 + 11 1.00 0.14 0.25 14 + 12 0.75 0.41 0.53 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.60 0.68 1404 + macro avg 0.68 0.36 0.43 1404 +weighted avg 0.78 0.60 0.65 1404 + samples avg 0.80 0.66 0.68 1404 diff --git a/games_march2025_cleaned_10k/MultinomialNB.txt b/games_march2025_cleaned_10k/MultinomialNB.txt new file mode 100644 index 0000000..bc74cf3 --- /dev/null +++ b/games_march2025_cleaned_10k/MultinomialNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.64 0.99 0.78 300 + 1 0.85 0.24 0.37 216 + 2 0.60 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 0.80 0.05 0.09 83 + 5 0.00 0.00 0.00 0 + 6 0.78 0.80 0.79 245 + 7 0.40 0.05 0.09 42 + 8 1.00 0.04 0.08 127 + 9 0.00 0.00 0.00 12 + 10 0.20 0.01 0.02 127 + 11 0.00 0.00 0.00 14 + 12 1.00 0.05 0.09 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.69 0.40 0.51 1404 + macro avg 0.45 0.16 0.17 1404 +weighted avg 0.68 0.40 0.39 1404 + samples avg 0.70 0.44 0.50 1404 diff --git a/games_march2025_cleaned_10k/RandomForestClassifier.txt b/games_march2025_cleaned_10k/RandomForestClassifier.txt new file mode 100644 index 0000000..6fbe546 --- /dev/null +++ b/games_march2025_cleaned_10k/RandomForestClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.80 0.88 0.84 300 + 1 0.78 0.55 0.64 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.06 0.11 83 + 5 0.00 0.00 0.00 0 + 6 0.74 0.78 0.76 245 + 7 0.00 0.00 0.00 42 + 8 0.84 0.24 0.38 127 + 9 0.00 0.00 0.00 12 + 10 0.91 0.24 0.38 127 + 11 1.00 0.14 0.25 14 + 12 1.00 0.25 0.39 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.48 0.59 1404 + macro avg 0.58 0.23 0.27 1404 +weighted avg 0.78 0.48 0.52 1404 + samples avg 0.77 0.54 0.60 1404 diff --git a/games_march2025_cleaned_10k/SVC-RBF-i10000.txt b/games_march2025_cleaned_10k/SVC-RBF-i10000.txt new file mode 100644 index 0000000..ff0c7b7 --- /dev/null +++ b/games_march2025_cleaned_10k/SVC-RBF-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.81 0.90 0.85 300 + 1 0.76 0.63 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.05 0.09 83 + 5 0.00 0.00 0.00 0 + 6 0.77 0.83 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.84 0.40 0.54 127 + 9 1.00 0.17 0.29 12 + 10 0.90 0.34 0.49 127 + 11 1.00 0.14 0.25 14 + 12 0.92 0.21 0.34 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.80 0.53 0.63 1404 + macro avg 0.64 0.26 0.32 1404 +weighted avg 0.79 0.53 0.56 1404 + samples avg 0.79 0.59 0.63 1404 diff --git a/games_march2025_cleaned_2k.csv b/games_march2025_cleaned_2k.csv new file mode 100644 index 0000000..806e982 --- /dev/null +++ b/games_march2025_cleaned_2k.csv @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:75ba38404995149bcb8e5a321459f73b4adf58597f85bab396dd054cc78c145d +size 15455174 diff --git a/games_march2025_cleaned_2k/BernoulliNB.txt b/games_march2025_cleaned_2k/BernoulliNB.txt new file mode 100644 index 0000000..f2237d4 --- /dev/null +++ b/games_march2025_cleaned_2k/BernoulliNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.75 0.90 0.82 300 + 1 0.72 0.68 0.70 216 + 2 0.50 0.08 0.14 86 + 3 0.27 0.07 0.11 46 + 4 0.40 0.07 0.12 83 + 5 0.00 0.00 0.00 0 + 6 0.77 0.82 0.79 245 + 7 0.33 0.10 0.15 42 + 8 0.67 0.40 0.50 127 + 9 0.00 0.00 0.00 12 + 10 0.71 0.37 0.49 127 + 11 0.00 0.00 0.00 14 + 12 0.49 0.31 0.38 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.70 0.55 0.62 1404 + macro avg 0.40 0.27 0.30 1404 +weighted avg 0.64 0.55 0.56 1404 + samples avg 0.73 0.59 0.61 1404 diff --git a/games_march2025_cleaned_2k/DecisionTreeClassifier.txt b/games_march2025_cleaned_2k/DecisionTreeClassifier.txt new file mode 100644 index 0000000..900c256 --- /dev/null +++ b/games_march2025_cleaned_2k/DecisionTreeClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.76 0.73 0.75 300 + 1 0.56 0.53 0.54 216 + 2 0.36 0.33 0.34 86 + 3 0.33 0.26 0.29 46 + 4 0.40 0.46 0.43 83 + 5 0.00 0.00 0.00 0 + 6 0.65 0.61 0.63 245 + 7 0.39 0.40 0.40 42 + 8 0.59 0.57 0.58 127 + 9 0.60 0.25 0.35 12 + 10 0.56 0.51 0.53 127 + 11 0.39 0.50 0.44 14 + 12 0.52 0.49 0.50 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.58 0.55 0.57 1404 + macro avg 0.44 0.40 0.41 1404 +weighted avg 0.58 0.55 0.57 1404 + samples avg 0.59 0.59 0.55 1404 diff --git a/games_march2025_cleaned_2k/GaussianNB.txt b/games_march2025_cleaned_2k/GaussianNB.txt new file mode 100644 index 0000000..83d7a2e --- /dev/null +++ b/games_march2025_cleaned_2k/GaussianNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.76 0.80 0.78 300 + 1 0.62 0.51 0.56 216 + 2 0.63 0.14 0.23 86 + 3 0.17 0.02 0.04 46 + 4 0.42 0.10 0.16 83 + 5 0.00 0.00 0.00 0 + 6 0.68 0.66 0.67 245 + 7 0.56 0.12 0.20 42 + 8 0.55 0.33 0.41 127 + 9 0.67 0.17 0.27 12 + 10 0.65 0.31 0.42 127 + 11 1.00 0.14 0.25 14 + 12 0.53 0.29 0.38 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.66 0.47 0.55 1404 + macro avg 0.52 0.26 0.31 1404 +weighted avg 0.62 0.47 0.51 1404 + samples avg 0.67 0.53 0.55 1404 diff --git a/games_march2025_cleaned_2k/GradientBoostingClassifier.txt b/games_march2025_cleaned_2k/GradientBoostingClassifier.txt new file mode 100644 index 0000000..7c8ce6e --- /dev/null +++ b/games_march2025_cleaned_2k/GradientBoostingClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.85 0.80 0.83 300 + 1 0.77 0.61 0.68 216 + 2 0.55 0.13 0.21 86 + 3 0.42 0.11 0.17 46 + 4 0.68 0.33 0.44 83 + 5 0.00 0.00 0.00 0 + 6 0.71 0.76 0.74 245 + 7 0.61 0.26 0.37 42 + 8 0.81 0.50 0.61 127 + 9 0.75 0.25 0.38 12 + 10 0.81 0.54 0.65 127 + 11 0.40 0.43 0.41 14 + 12 0.69 0.42 0.53 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.76 0.57 0.65 1404 + macro avg 0.57 0.37 0.43 1404 +weighted avg 0.74 0.57 0.63 1404 + samples avg 0.76 0.63 0.65 1404 diff --git a/games_march2025_cleaned_2k/LinearSVC-i5000.txt b/games_march2025_cleaned_2k/LinearSVC-i5000.txt new file mode 100644 index 0000000..df82b40 --- /dev/null +++ b/games_march2025_cleaned_2k/LinearSVC-i5000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.85 0.87 0.86 300 + 1 0.76 0.66 0.70 216 + 2 0.77 0.20 0.31 86 + 3 0.00 0.00 0.00 46 + 4 0.76 0.27 0.39 83 + 5 0.00 0.00 0.00 0 + 6 0.78 0.81 0.79 245 + 7 0.89 0.19 0.31 42 + 8 0.77 0.60 0.67 127 + 9 1.00 0.58 0.74 12 + 10 0.85 0.54 0.66 127 + 11 1.00 0.29 0.44 14 + 12 0.82 0.42 0.56 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.80 0.61 0.69 1404 + macro avg 0.66 0.39 0.46 1404 +weighted avg 0.78 0.61 0.66 1404 + samples avg 0.81 0.67 0.69 1404 diff --git a/games_march2025_cleaned_2k/LogisticRegression-i1000.txt b/games_march2025_cleaned_2k/LogisticRegression-i1000.txt new file mode 100644 index 0000000..b7926d4 --- /dev/null +++ b/games_march2025_cleaned_2k/LogisticRegression-i1000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.78 0.91 0.84 300 + 1 0.78 0.62 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.04 0.07 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.81 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.90 0.34 0.49 127 + 9 0.00 0.00 0.00 12 + 10 0.89 0.25 0.39 127 + 11 0.00 0.00 0.00 14 + 12 0.88 0.14 0.24 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.50 0.61 1404 + macro avg 0.50 0.22 0.26 1404 +weighted avg 0.77 0.50 0.53 1404 + samples avg 0.77 0.56 0.60 1404 diff --git a/games_march2025_cleaned_2k/LogisticRegression-i10000.txt b/games_march2025_cleaned_2k/LogisticRegression-i10000.txt new file mode 100644 index 0000000..b7926d4 --- /dev/null +++ b/games_march2025_cleaned_2k/LogisticRegression-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.78 0.91 0.84 300 + 1 0.78 0.62 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.04 0.07 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.81 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.90 0.34 0.49 127 + 9 0.00 0.00 0.00 12 + 10 0.89 0.25 0.39 127 + 11 0.00 0.00 0.00 14 + 12 0.88 0.14 0.24 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.50 0.61 1404 + macro avg 0.50 0.22 0.26 1404 +weighted avg 0.77 0.50 0.53 1404 + samples avg 0.77 0.56 0.60 1404 diff --git a/games_march2025_cleaned_2k/MLPClassifier-i10000.txt b/games_march2025_cleaned_2k/MLPClassifier-i10000.txt new file mode 100644 index 0000000..c4634dc --- /dev/null +++ b/games_march2025_cleaned_2k/MLPClassifier-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.84 0.85 0.84 300 + 1 0.73 0.67 0.70 216 + 2 0.74 0.30 0.43 86 + 3 0.50 0.02 0.04 46 + 4 0.69 0.24 0.36 83 + 5 0.00 0.00 0.00 0 + 6 0.79 0.79 0.79 245 + 7 0.86 0.14 0.24 42 + 8 0.76 0.63 0.69 127 + 9 1.00 0.33 0.50 12 + 10 0.81 0.52 0.63 127 + 11 1.00 0.14 0.25 14 + 12 0.75 0.41 0.53 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.60 0.68 1404 + macro avg 0.68 0.36 0.43 1404 +weighted avg 0.78 0.60 0.65 1404 + samples avg 0.80 0.66 0.68 1404 diff --git a/games_march2025_cleaned_2k/MultinomialNB.txt b/games_march2025_cleaned_2k/MultinomialNB.txt new file mode 100644 index 0000000..bc74cf3 --- /dev/null +++ b/games_march2025_cleaned_2k/MultinomialNB.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.64 0.99 0.78 300 + 1 0.85 0.24 0.37 216 + 2 0.60 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 0.80 0.05 0.09 83 + 5 0.00 0.00 0.00 0 + 6 0.78 0.80 0.79 245 + 7 0.40 0.05 0.09 42 + 8 1.00 0.04 0.08 127 + 9 0.00 0.00 0.00 12 + 10 0.20 0.01 0.02 127 + 11 0.00 0.00 0.00 14 + 12 1.00 0.05 0.09 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.69 0.40 0.51 1404 + macro avg 0.45 0.16 0.17 1404 +weighted avg 0.68 0.40 0.39 1404 + samples avg 0.70 0.44 0.50 1404 diff --git a/games_march2025_cleaned_2k/RandomForestClassifier.txt b/games_march2025_cleaned_2k/RandomForestClassifier.txt new file mode 100644 index 0000000..6fbe546 --- /dev/null +++ b/games_march2025_cleaned_2k/RandomForestClassifier.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.80 0.88 0.84 300 + 1 0.78 0.55 0.64 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.06 0.11 83 + 5 0.00 0.00 0.00 0 + 6 0.74 0.78 0.76 245 + 7 0.00 0.00 0.00 42 + 8 0.84 0.24 0.38 127 + 9 0.00 0.00 0.00 12 + 10 0.91 0.24 0.38 127 + 11 1.00 0.14 0.25 14 + 12 1.00 0.25 0.39 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.79 0.48 0.59 1404 + macro avg 0.58 0.23 0.27 1404 +weighted avg 0.78 0.48 0.52 1404 + samples avg 0.77 0.54 0.60 1404 diff --git a/games_march2025_cleaned_2k/SVC-RBF-i10000.txt b/games_march2025_cleaned_2k/SVC-RBF-i10000.txt new file mode 100644 index 0000000..ff0c7b7 --- /dev/null +++ b/games_march2025_cleaned_2k/SVC-RBF-i10000.txt @@ -0,0 +1,21 @@ + precision recall f1-score support + + 0 0.81 0.90 0.85 300 + 1 0.76 0.63 0.69 216 + 2 1.00 0.03 0.07 86 + 3 0.00 0.00 0.00 46 + 4 1.00 0.05 0.09 83 + 5 0.00 0.00 0.00 0 + 6 0.77 0.83 0.80 245 + 7 0.00 0.00 0.00 42 + 8 0.84 0.40 0.54 127 + 9 1.00 0.17 0.29 12 + 10 0.90 0.34 0.49 127 + 11 1.00 0.14 0.25 14 + 12 0.92 0.21 0.34 106 + 13 0.00 0.00 0.00 0 + + micro avg 0.80 0.53 0.63 1404 + macro avg 0.64 0.26 0.32 1404 +weighted avg 0.79 0.53 0.56 1404 + samples avg 0.79 0.59 0.63 1404 diff --git a/notebook.ipynb b/notebook.ipynb new file mode 100644 index 0000000..3307ceb --- /dev/null +++ b/notebook.ipynb @@ -0,0 +1,530 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "a3a7634f", + "metadata": {}, + "source": [ + "# Machine Learning project in SoSe 2025 at HTW Saar\n", + "## Idea\n", + "The goal of this project is predicting the genre(s) of a game/bundle through its given description(s)\n", + "\n", + "## Dataset\n", + "For our project we use a Steam Dataset provided on moodle, since it has all information we plan on using.\n", + "The Dataset has been cut to only 2000 data points to be runnable on weaker devices." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "3116b75f", + "metadata": { + "jupyter": { + "is_executing": true + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " appid name release_date required_age price dlc_count \\\n", + "0 730 Counter-Strike 2 2012-08-21 0 0.0 1 \n", + "\n", + " detailed_description \\\n", + "0 For over two decades, Counter-Strike has offer... \n", + "\n", + " about_the_game \\\n", + "0 For over two decades, Counter-Strike has offer... \n", + "\n", + " short_description reviews ... \\\n", + "0 For over two decades, Counter-Strike has offer... NaN ... \n", + "\n", + " average_playtime_2weeks median_playtime_forever median_playtime_2weeks \\\n", + "0 879 5174 350 \n", + "\n", + " discount peak_ccu tags \\\n", + "0 0 1212356 {'FPS': 90857, 'Shooter': 65397, 'Multiplayer'... \n", + "\n", + " pct_pos_total num_reviews_total pct_pos_recent num_reviews_recent \n", + "0 86 8632939 82 96473 \n", + "\n", + "[1 rows x 47 columns]\n" + ] + } + ], + "source": [ + "import numpy as np\n", + "import pandas as pd\n", + "from sklearn import set_config\n", + "\n", + "set_config(transform_output=\"pandas\")\n", + "\n", + "dataset = pd.read_csv(\"./games_march2025_cleaned_2k.csv\",sep=\",\")\n", + "print(dataset.head(1))" + ] + }, + { + "cell_type": "markdown", + "id": "cba9750a", + "metadata": {}, + "source": [ + "## Preparation of the Dataset\n", + "### Removing Uniques\n", + "We would remove the following features from the Training-Set as they can/could uniquely identify a datapoint, but we don't as they will be removed in the next step anyway\n", + "- AppId\n", + "- Name of the Game\n", + "- Realease Date\n", + "- Reviews\n", + "- Header Image\n", + "- Website\n", + "- Support URL\n", + "- Support Email\n", + "- MetaCritic URL\n", + "- Developer\n", + "- Publisher\n", + "- Screenshots\n", + "- Movies\n", + "- Estimated Owners" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "d159117377f3633c", + "metadata": {}, + "outputs": [], + "source": [ + "#dataset.drop(['appid', 'name', 'release_date', 'reviews', 'header_image', 'website', 'support_url', 'support_email', 'metacritic_url', 'notes', 'developers', 'publishers', 'screenshots', 'movies', 'estimated_owners'], axis=1, inplace=True)\n", + "#print(dataset.head())" + ] + }, + { + "cell_type": "markdown", + "id": "e1b28ddd69f1e9a6", + "metadata": {}, + "source": [ + "## Hold onto necessary information\n", + "Our model should turn a textual description of a game into its genre. For that we need all the textual information a game has, as well as the genres of the game.\n", + "We use a ColumnTransformer to drop all unnecessary lines, merge all descriptions of a game into one big description and hold onto the genres\n", + "\n", + "It is important to use ``verbose_feature_names_out=False`` so the feature names don't get changed" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "986fbb31a7ae0d8b", + "metadata": { + "jupyter": { + "is_executing": true + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " desc \\\n", + "0 For over two decades, Counter-Strike has offer... \n", + "1 LAND, LOOT, SURVIVE! Play PUBG: BATTLEGROUNDS ... \n", + "2 The most-played game on Steam. Every day, mill... \n", + "3 When a young street hustler, a retired bank ro... \n", + "4 Edition Comparison Ultimate Edition The Tom Cl... \n", + "\n", + " genres \n", + "0 ['Action', 'Free To Play'] \n", + "1 ['Action', 'Adventure', 'Massively Multiplayer... \n", + "2 ['Action', 'Strategy', 'Free To Play'] \n", + "3 ['Action', 'Adventure'] \n", + "4 ['Action'] \n" + ] + } + ], + "source": [ + "from sklearn.compose import ColumnTransformer\n", + "from sklearn.preprocessing import FunctionTransformer\n", + "\n", + "# desc, genres\n", + "column_transformer = ColumnTransformer([\n", + " # merge all descriptions\n", + " ('desc', FunctionTransformer(lambda X: X.fillna('').agg(' '.join, axis=1).to_frame(name=\"desc\")),\n", + " ['detailed_description', 'about_the_game', 'short_description']),\n", + " ('pass', 'passthrough', ['genres']),\n", + " ],\n", + " verbose_feature_names_out=False\n", + ")\n", + "dataset = column_transformer.fit_transform(dataset)\n", + "print(dataset.head())" + ] + }, + { + "cell_type": "markdown", + "id": "f9b89c0645811564", + "metadata": {}, + "source": [ + "### Adding missing Information\n", + "Some Games might not have any descriptions. For these we Input an Empty String\n", + "**TODO: check if dropna and fillna numeric_only is needed, as we dont have any numbers**" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "44239f6b7fd23cde", + "metadata": {}, + "outputs": [], + "source": [ + "# missing numeric values => mean\n", + "dataset.fillna(dataset.mean(numeric_only=True), inplace=True)\n", + "# missing strings => empty string?\n", + "dataset.fillna('', inplace=True)\n", + "# drop all lines with missing values\n", + "dataset.dropna(inplace=True)" + ] + }, + { + "cell_type": "markdown", + "id": "ca5b59b9fa8160a0", + "metadata": {}, + "source": [ + "## Transform Genres\n", + "The genre information currently is a string holding a python array of genres. While this is machine-readable, we need One-Hot-Encoding for our model to work.\n", + "\n", + "#### Serializing the String-Array\n", + "The \"ast\" library can interpret python strings as python code, and as such will be used for serializing the genres." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ebc5a24e9bc87fdd", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "0 [Action, Free To Play]\n", + "1 [Action, Adventure, Massively Multiplayer, Fre...\n", + "2 [Action, Strategy, Free To Play]\n", + "3 [Action, Adventure]\n", + "4 [Action]\n", + "Name: genres, dtype: object\n" + ] + } + ], + "source": [ + "import ast\n", + "\n", + "dataset['genres'] = dataset['genres'].map(lambda s: ast.literal_eval(s))\n", + "print(dataset['genres'].head())" + ] + }, + { + "cell_type": "markdown", + "id": "f90756f9ad9211f4", + "metadata": {}, + "source": [ + "#### One-Hot-Encoding an Python-Array\n", + "The sklearn ``OneHotEncoder()`` is only able to work with an 1D Array of different classes, such as ``['Politics', 'Sport', 'Culture']``. Every datapoint can only have one concurrent classification.\n", + "Steam allows an app/bundle to have multiple genres. As such, our dataset has an 2D Array of different classes, which sklearn's ``MultiLabelBinarizer()`` does support." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "d2c3527a5fc876bf", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " Action Adventure Casual Early Access Free To Play Gore Indie \\\n", + "0 1 0 0 0 1 0 0 \n", + "1 1 1 0 0 1 0 0 \n", + "2 1 0 0 0 1 0 0 \n", + "3 1 1 0 0 0 0 0 \n", + "4 1 0 0 0 0 0 0 \n", + "\n", + " Massively Multiplayer RPG Racing Simulation Sports Strategy Violent \n", + "0 0 0 0 0 0 0 0 \n", + "1 1 0 0 0 0 0 0 \n", + "2 0 0 0 0 0 1 0 \n", + "3 0 0 0 0 0 0 0 \n", + "4 0 0 0 0 0 0 0 \n" + ] + } + ], + "source": [ + "from sklearn.preprocessing import MultiLabelBinarizer\n", + "\n", + "mlb_genres = MultiLabelBinarizer()\n", + "genres_encoded = mlb_genres.fit_transform(dataset.pop('genres'))\n", + "genres_df = pd.DataFrame(genres_encoded, columns=mlb_genres.classes_)\n", + "print(genres_df.head())" + ] + }, + { + "cell_type": "markdown", + "id": "671c01f9f4ae66d9", + "metadata": {}, + "source": [ + "With this, our target matrix is completed." + ] + }, + { + "cell_type": "markdown", + "id": "f5436c87", + "metadata": {}, + "source": [ + "### Structurizing Text\n", + "If we want our Model to be able to use text as an input, we have to vectorize the text. TF-IDF (Inverse Document Frequency) is an easy way of transforming each word into a feature with a 0 to 1 value. **TODO: filter out stopwords**" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "4e8b407c", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " 00 000 000km 000th 00am 00f 00i 00p 00v 01 ... 이터널 이터널리턴 \\\n", + "0 0.0 0.0 0.0 0.00000 0.0 0.0 0.0 0.0 0.0 0.0 ... 0.0 0.0 \n", + "1 0.0 0.0 0.0 0.00000 0.0 0.0 0.0 0.0 0.0 0.0 ... 0.0 0.0 \n", + "2 0.0 0.0 0.0 0.14649 0.0 0.0 0.0 0.0 0.0 0.0 ... 0.0 0.0 \n", + "3 0.0 0.0 0.0 0.00000 0.0 0.0 0.0 0.0 0.0 0.0 ... 0.0 0.0 \n", + "4 0.0 0.0 0.0 0.00000 0.0 0.0 0.0 0.0 0.0 0.0 ... 0.0 0.0 \n", + "\n", + " 이현준 정대찬 중입니다 철권 토탈워 페르소나 한국어 한글을 \n", + "0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", + "1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", + "2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", + "3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", + "4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 \n", + "\n", + "[5 rows x 29351 columns]\n" + ] + } + ], + "source": [ + "from sklearn.feature_extraction.text import TfidfVectorizer\n", + "\n", + "vectorizer = TfidfVectorizer()\n", + "tfidf_matrix = vectorizer.fit_transform(dataset['desc']) # matrix, not pandas df\n", + "tfidf_df = pd.DataFrame(tfidf_matrix.toarray(), columns=vectorizer.get_feature_names_out())\n", + "print(tfidf_df.head())" + ] + }, + { + "cell_type": "markdown", + "id": "ad84e777", + "metadata": {}, + "source": [ + "With this our feature matrix is completed" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "86d9da42f4df8e49", + "metadata": {}, + "outputs": [], + "source": [ + "X = tfidf_df\n", + "y = genres_df" + ] + }, + { + "cell_type": "markdown", + "id": "aeb782668f311cd8", + "metadata": {}, + "source": [ + "## The Model\n", + "\n", + "#### Removing unpredicatble Datapoints\n", + "Some Datapoints don't have a genre assigned (all feature values in y are 0). The model we use can't handle such cases, thus they have to be removed.\n", + "We filter after all values that we can use with a mask, and apply that mask to our matrices." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "4919bf1b37d171a7", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "13\n" + ] + } + ], + "source": [ + "mask = y.sum(axis=1).map(lambda x: x > 0)\n", + "print((mask == False).sum()) # count of unpredictable datapoints\n", + "\n", + "X_clean = X[mask]\n", + "y_clean = y[mask]" + ] + }, + { + "cell_type": "markdown", + "id": "091d7e13", + "metadata": {}, + "source": [ + "# Splitting up data\n", + "We have to split up our data into training and testing data.\n", + "Using random_state=0 guarantees reproducability." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "cfbf3787", + "metadata": { + "jupyter": { + "is_executing": true + } + }, + "outputs": [], + "source": [ + "from sklearn.model_selection import train_test_split\n", + "\n", + "X_train, X_test, y_train, y_test = train_test_split(X_clean, y_clean, random_state=0)" + ] + }, + { + "cell_type": "markdown", + "id": "12b5283d", + "metadata": {}, + "source": [ + "# Model Selection\n", + "**TODO Deciding which model to use for this task**\n", + "\n", + "As a game can have multiple genres, our Model(s) has to be capable of multi-label-classification. sklearn's ``MultiOutputClassifier`` can do this. As a backend for ``MultiOutputClassifier`` we use ``LogisticRegression``" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8c1d72c4532bd509", + "metadata": {}, + "outputs": [], + "source": [ + "from sklearn.linear_model import LogisticRegression\n", + "from sklearn.multioutput import MultiOutputClassifier\n", + "\n", + "# n_jobs=1 since there seems to be some multithreading join issue in sklearn (or my pc is to bad)\n", + "multi_target_clf = MultiOutputClassifier(LogisticRegression(max_iter=1337, random_state=0), n_jobs=1)\n", + "\n", + "multi_target_clf.fit(X_train, y_train)\n", + "\n", + "y_pred = multi_target_clf.predict(X_test)" + ] + }, + { + "cell_type": "markdown", + "id": "0faa9856", + "metadata": {}, + "source": [ + "# Evaluation\n", + "**TODO Test the Model with the test data**" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e2ebea6945193e07", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " precision recall f1-score support\n", + "\n", + " 0 0.78 0.91 0.84 300\n", + " 1 0.78 0.62 0.69 216\n", + " 2 1.00 0.03 0.07 86\n", + " 3 0.00 0.00 0.00 46\n", + " 4 1.00 0.04 0.07 83\n", + " 5 0.00 0.00 0.00 0\n", + " 6 0.79 0.81 0.80 245\n", + " 7 0.00 0.00 0.00 42\n", + " 8 0.90 0.34 0.49 127\n", + " 9 0.00 0.00 0.00 12\n", + " 10 0.89 0.25 0.39 127\n", + " 11 0.00 0.00 0.00 14\n", + " 12 0.88 0.14 0.24 106\n", + " 13 0.00 0.00 0.00 0\n", + "\n", + " micro avg 0.79 0.50 0.61 1404\n", + " macro avg 0.50 0.22 0.26 1404\n", + "weighted avg 0.77 0.50 0.53 1404\n", + " samples avg 0.77 0.56 0.60 1404\n", + "\n" + ] + } + ], + "source": [ + "from sklearn.metrics import classification_report\n", + "\n", + "print(classification_report(y_test, y_pred, zero_division=0.0))" + ] + }, + { + "cell_type": "markdown", + "id": "2aeb6fc2", + "metadata": {}, + "source": [ + "# Optimization\n", + "**TODO optimize the model based on the test results**" + ] + }, + { + "cell_type": "markdown", + "id": "79b20645", + "metadata": {}, + "source": [ + "# Validation\n", + "**TODO Predict actual values**" + ] + }, + { + "cell_type": "markdown", + "id": "3b709fb7", + "metadata": {}, + "source": [ + "# Conclusion and outlook\n", + "**TODO Write a conclusion and outlook what can be done and where the issues were.**" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.13.3" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/test_script.py b/test_script.py new file mode 100644 index 0000000..de7e833 --- /dev/null +++ b/test_script.py @@ -0,0 +1,133 @@ + + +#### INITIALIZE + +import numpy as np +import pandas as pd +from sklearn import set_config +set_config(transform_output="pandas") # dataframe supremacy + +# load data +# appid,name,release_date,required_age,price,dlc_count,detailed_description,about_the_game,short_description,reviews,header_image,website,support_url,support_email,windows,mac,linux,metacritic_score,metacritic_url,achievements,recommendations,notes,supported_languages,full_audio_languages,packages,developers,publishers,categories,genres,screenshots,movies,user_score,score_rank,positive,negative,estimated_owners,average_playtime_forever,average_playtime_2weeks,median_playtime_forever,median_playtime_2weeks,discount,peak_ccu,tags,pct_pos_total,num_reviews_total,pct_pos_recent,num_reviews_recent +dataset = pd.read_csv("./games_march2025_cleaned_2k.csv",sep=",") +print(dataset.head()) + + + + +#### DROP UNIQUES +print("DROP") + +#TODO: wird eh unten beim transformer deleted + +# appid,name,release_date,required_age,price,dlc_count,detailed_description,about_the_game,short_description,reviews,header_image,website,support_url,support_email,windows,mac,linux,metacritic_score,metacritic_url,achievements,recommendations,notes,supported_languages,full_audio_languages,packages,developers,publishers,categories,genres,screenshots,movies,user_score,score_rank,positive,negative,estimated_owners,average_playtime_forever,average_playtime_2weeks,median_playtime_forever,median_playtime_2weeks,discount,peak_ccu,tags,pct_pos_total,num_reviews_total,pct_pos_recent,num_reviews_recent +#dataset.drop(['appid', 'name', 'release_date', 'reviews', 'header_image', 'website', 'support_url', 'support_email', +# 'metacritic_url', 'notes', 'developers', 'publishers', 'screenshots', 'movies', 'estimated_owners'], +# axis=1, inplace=True) +#print(dataset.head()) + +#### STRUCTURIZE AND STANDARDIZE +print("STRUCTURE") + +from sklearn.compose import ColumnTransformer +from sklearn.preprocessing import FunctionTransformer + + +# desc, genres, tags +column_transformer = ColumnTransformer([ + # merge all descriptions + ('desc', FunctionTransformer(lambda X: X.fillna('').agg(' '.join, axis=1).to_frame(name="desc")), + ['detailed_description', 'about_the_game', 'short_description']), + # genre -> actual genre, but very coarse + # tags -> user defined tags; title num list + #TODO: decide whether we drop tags + ('pass', 'passthrough', ['genres']),#, 'tags' + ], + verbose_feature_names_out=False +) +dataset = column_transformer.fit_transform(dataset) +print(dataset) + + + +#### SET MISSING VALUES +print("SETMISS") + + +# Setting missing numeric values to the mean +dataset.fillna(dataset.mean(numeric_only=True), inplace=True) +# Setting missing text values to 'Unknown' +dataset.fillna('', inplace=True) +# Setting missing values in other columns to NaN +dataset.dropna(inplace=True) + + + + +##### STRUCTURIZE GENRES to onehot +from sklearn.preprocessing import MultiLabelBinarizer +import ast +#serialize array +dataset['genres'] = dataset['genres'].map(lambda s: ast.literal_eval(s)) +print(dataset['genres']) # in py but not yet onehotenc + +# MultiLabelBinarizer does onehotenc for arrays +mlb_genres = MultiLabelBinarizer() +genres_encoded = mlb_genres.fit_transform(dataset.pop('genres')) +genres_count = len(mlb_genres.classes_) # for multi-label classifiction later + +genres_df = pd.DataFrame(genres_encoded, columns=mlb_genres.classes_) +print(genres_df) +#dataset = pd.concat([dataset, genres_df], axis=1) +#print(dataset) + + +#### convert text to bag of words + +## Count vs Tfidf vectorizer +from sklearn.feature_extraction.text import TfidfVectorizer +vectorizer = TfidfVectorizer() +tfidf_matrix = vectorizer.fit_transform(dataset['desc']) # matrix +tfidf_df = pd.DataFrame(tfidf_matrix.toarray(), columns=vectorizer.get_feature_names_out()) +print(tfidf_df) + + +##### MODEL +print("MODEL") + +from sklearn.linear_model import LogisticRegression +from sklearn.multioutput import MultiOutputClassifier +from sklearn.metrics import classification_report + + +X = tfidf_df +y = genres_df + + +# cleanup datapoints that dont have a target value (all target columns are 0) +mask = y.sum(axis=1).map(lambda x: x > 0) +#print((mask == False).sum()) #31 cases with all target columns 0 +X_clean = X[mask] +y_clean = y[mask] + +# Split dataset +from sklearn.model_selection import train_test_split +X_train, X_test, y_train, y_test = train_test_split(X_clean, y_clean, random_state=0) + + +# we want to have multiple possible outputs (multi-label-classficiation) -> multioutputclassifier +# logi regression is our base system +# n_jobs=1 since there seems to be some multithreading join issue in sklearn (or my pc is too bad) +multi_target_clf = MultiOutputClassifier(LogisticRegression(max_iter=1337, random_state=0), n_jobs=1) + +# model training +multi_target_clf.fit(X_train, y_train) + +# predict against test data +y_pred = multi_target_clf.predict(X_test) + +# print prec, recall, f1 etc +print(classification_report(y_test, y_pred, zero_division=0.0)) + + +#print(f"Trainingsdaten: {X_train.shape}, Testdaten: {X_test.shape}")