Machine Learning

Photo by tungnguyen0905 | Pixabay

List of my Sample Projects on Machine Learning

Click on Project Name to View the Project

• Classification of Iris Plants (Python Tensorflow): The project aims to train a Multi-Layer Perceptron Machine Learning model for classification using Tensorflow package. Using four features such as sepal length, sepal width, petal length, and petal width), the model should classify the observation whether it is an iris setosa, iris virginica, or iris versicolor.
• Predicting Real Estate Prices (Python): Sindian District New Taipei City is an urbanized city in Taiwan. In this project the explanatory variables considered to estimate house price include transaction date, house age, distance from the nearest Mass Rapid Transit (MRT), number of convenience stores nearby, and the latitude and longitude of the property.
• Classification of Exercise Movements (R): In this project we will explore the exercise movements data set, train machine learning models (using Classification Trees, Gradient Boost, and Random Forest), assess its respective accuracy in classifying these exercise movements using the validation set, and predict classifications made by the better performing model using the test set.   
• Classification on Wine Data using KNN (Python): The team wishes to learn a model that will classify wines. This project will utilize the K-Nearest Neighbor models with different distance. The KNN models will be assessed in terms accuracy in classifying the wines at the same time speed in processing.   
• KMeans++ for Mall Customer Insight (Python): This uses Pyspark for clustering customers using Kmeans++. The administration is looking to cluster their customers in terms of features such as sex, age, annual income (in thousand), and spending score. This is to be used in making decisions for marketing and promotions purposes.   
• Predicting Credit Card Fraud (Python):The company is interested to implement a machine learning model that will identify fraudulent transactions. This Project uses Logistic Regression to classify credit card transactions. The data set provided contains 30 features and 284,807 records. The V-Columns represent components from Principal Component Analysis, the variable labeled Amount unchanged, the variable Class is although numerical (ready for modelling) has 1 which refers to fraudulent transaction whereas 0 is considered non-fraudulent.   
• Predicting High Risk Credit Card Applicants (R):This case presents a model for financial institutions to predict high risk credit card customers based on lifestyle variables. It uses the support vector machine model to check whether applicant is a considered high risk credit card customer or not. It may serve as reference for approval of credit card applications.