This repo contains the projects for Udacity Machine Learning Engineering Nanodegree.
Created a simple decision algorithms and increased its complexity until it could accurately predict the outcomes for at least 80% of the passengers in the provided data.
Keywords: Python, NumPy, Pandas, Matplotlib, simple algorithms
Built a regression model to predict the value of a given house in the Boston real estate market using various statistical analysis tools. Identified the best price that a client can sell their house utilizing machine learning algorithms.
Keywords: statistical analysis, metric performance, cross validation, bias/variance tradeoffs, learning curves, model complexity
Investigated factors that affect the likelihood of charity donations being made based on real census data. Developed a naive classifier to compare testing results to. Trained and tested several supervised machine learning models on preprocessed census data to predict the likelihood of donations. Selected the best model based on accuracy, a modified F-scoring metric, and algorithm efficiency.
Keywords: classification, logistic regression, decision trees, ensemble methods, model tuning
Reviewed unstructured data to understand the patterns and natural categories that the data fits into. Used multiple algorithms and both empirically and theoretically compared and contrasted their results. Made predictions about the natural categories of multiple types in a dataset, then checked these predictions against the result of unsupervised analysis.
Keywords: clustering, pca, feature selection, k means, gaussian mixture model
Applied reinforcement learning to build a simulated vehicle navigation agent. Identified the environment the agent operates in and the possible states the agent can be in. Implemented and optimized a Q-Learning algorithm to allow the agent to automatically learn an optimal driving strategy based on rewards and penalties.
Keywords: reinforcement learning, q-learning
Classified images from the CIFAR-10 dataset. The dataset was preprocessed (image normalization, label one-hot encoding), then trained a convolutional neural network with convolutional layer, max pool layer and fully connected layer on all the samples. The model was then evaluated on test samples.
Keywords: classification, convolutional neural network, tensorflow, keras, ResNet50
Capstone project for this Nanodegree. Built a stock price indicator that took daily trading data over a certain date range as input, and outputs projected estimates for given query dates. Created time-series cross validation method on stock data. Applied two metrics and three regression models.
Keywords: supervised learning, time series cross validation, linear regression, knn, svr, model refinement