Merge pull request #641 from Manav173/moviegenre

Added Movie Genre Classification [Issue No. #566]

Movie Genre Classification/Readme.md

@@ -0,0 +1,48 @@
+## Movie Genre Classification
+
+### Data Set
+The dataset used for this project is stored in the file `train_data.txt`.
+
+### Use
+The "Movie Genre.ipynb" file is a Jupyter notebook that analyzes movie genres using machine learning techniques. 
+
+### Dependencies
+The notebook imports the following libraries:
+- Matplotlib for plotting visualizations
+- Seaborn for enhancing the visualizations with different styles and color palettes
+- GridSearchCV from scikit-learn for hyperparameter tuning in the Logistic Regression model
+- LogisticRegression from scikit-learn for training the classification model
+
+## APPROACH
+The provided file is a Jupyter notebook named "Movie Genre.ipynb" that analyzes movie genres using machine learning techniques. Here is a summary of its contents:
+
+Data Preprocessing
+- The notebook starts by loading the movie data into a DataFrame named "train_df".
+- It extracts the unique genres from the 'Genre' column and displays the list of 28 unique genres
+- A horizontal bar plot is created to visualize the distribution of genres, showing the count of each genre
+- The 'Description' column is preprocessed using the 'preprocessing' function, and the cleaned descriptions are    added to the DataFrame as a new column 'Description_clean'
+
+Exploratory Data Analysis
+- The first few rows of the DataFrame are displayed to show the structure of the data
+- Another bar plot is created to visualize the distribution of genres using a different color palette and edge color
+
+Feature Engineering and Modeling
+- Unnecessary columns ('Title', 'Id', 'Description') are dropped from the DataFrame, leaving only 'Genre' and 'Description_clean'
+- The 'Genre' column is encoded using a dictionary mapping each genre to a unique integer value
+- The encoded genres are added to the DataFrame as a new column 'Genre_encoded'
+- The data is split into features (X) and target variable (y), where X is the 'Description_clean' column and y is the 'Genre_encoded' column
+- The text data in X is vectorized using TF-IDF vectorization with a maximum of 5000 features
+- A Logistic Regression model is trained using GridSearchCV to find the best hyperparameter (C) based on 5-fold cross-validation
+- The best model is evaluated on the test set, and the score is printed
+- A classification report is generated to show the precision, recall, f1-score, and support for each genre
+
+
+The notebook demonstrates the process of preprocessing movie data, exploring genre distributions, and training a Logistic Regression model to classify movies based on their descriptions. The model's performance is evaluated using the test set and a classification report.
+
+## Author
+
+### [Manav Malhotra](https://github.com/Manav173)
+
+## License
+
+This project is licensed under the MIT License.