Machine-Learning-Classification-Models

This repo covers the basic machine learning classification projects/problems using various machine learning classification techniques and MLP Neural Network through scikit learn library

[Project 1 : Gender Voice Classification ]

• The file cbndex1.ipynb was the very first attempt/beginner problem solution I worked on. It does basic LogisticRegression to classify voice by male or female and displays the Confusion Matrix, with Accuracy, Precision, and Recall Score to evaluate the classifier

• The file GenderVoice.ipynb is the latest edition, applying complex functions and creating a framework where data can go through few models and train test, predict, and evaluate which I learned as I progressed through the intermediate and advanced training and applied it to the very first classification problem Gender Voice : Using Logistic Regression Classifier, Naive Bayes Classifier, K-nearest-neighbors Classifier, K-nearest-neighbors Classifier, and Multilayer Preceptron Classifier, Decision Trees and Random Forest for warm start ML to classify the gender voice.

[Project 2 : German Credit Classification ]

• The file German_Credit.ipynb is the latest edition, applying complex functions and creating a framework where data can go through few models and train test, predict, and evaluate which I learned as I progressed through the intermediate and advanced training and applied it to this problem German Credit : Using Logistic Regression Classifier, Naive Bayes Classifier, K-nearest-neighbors Classifier, K-nearest-neighbors Classifier, and Multilayer Preceptron Classifier, Decision Trees and Random Forest for warm start ML to classify person by good or bad class in terms of credit, and evaluating the result by Classification Report.

[Project 3 : BBC Text Classification ]

• The file cltxtex3.ipynb was the very first attempt/beginner problem solution I worked on. In the exercise we covered different methods of feature extraction of text such has count vectorizer and TfidfVectorizer, then we proceeded to train and test the model using the Decision tree classifier with max depth of 10.

• The file BBC_text_classification.ipynb is the latest edition, applying complex functions and creating a framework where data can go through few models and train test, predict, and evaluate which I learned as I progressed through the intermediate and advanced training and applied it to the very first text classification problem BBC dataset: Using Linear Support Vector Classifier, decision tree, and Multilayer Preceptron Classifier to classify/catagorize the news articles.