ML_Practice

Brushing up on ML skills

NaiveBayesian.ipynb

• Introduction to Naive Bayes as a probabilistic classifier.
• Implemented a simple Naive Bayes classifier using categorical likelihoods.

LogisticRegression.ipynb

• Introduction to binary classification and the sigmoid function.
• Binary Cross-Entropy Loss: Overview of binary cross-entropy and its role in penalizing incorrect classifications.
• Gradient Descent: Implemented gradient descent to optimize weights in Logistic Regression.
• L2 Regularization: Added L2 regularization to prevent overfitting and explored the impact of regularization strength.

Churn Data/LogisticRegression.ipynb

• Using real-world churn data (https://www.kaggle.com/datasets/blastchar/telco-customer-churn).
• Encoding Categorical Data: Explored Label Encoding vs. One-Hot Encoding, and implemented both.
• Logistic Regression with Categorical Encoding: Compared the performance of one-hot vs. label-encoded data.
• Logistic Regression with class weights for unbalanced data

DecisionTrees.ipynb

• Using real-world churn data (https://www.kaggle.com/datasets/blastchar/telco-customer-churn)
• Label Encoding
• Implementing decision tree from scratch
• Cross Validation
• Using cProfile to see stats for different functions