Bayesian Probabilistic Programming & K-Clustering

Machine learning algorithms using supervised and unsupervised learning approaches are used to develop data analysis driven solutions. The Bayesian Machine Learning and Probabilistic program classifies flower characteristics and helps the user identify whether unlabeled data can be fit in the trained model. The K-Clustering Method sorts colours from an image via an iterative process

Data Analysis Results

Flower Type Distribution

Flower Length Distribution

Bayesian-Gaussian-Classification

My team mate and I developed a program that applies univariate Gaussian class condition probability density. The goal was to design a 2-class minimum-error classifier (dichotomizer) to classify IRIS samples into Iris Setosa or Iris Versicolour with respect to feature Sepal Width. In short, we classified flower features. The source code can be found on my Github.

Sorted Colour Results

Error Plot

K-Clustering

We applied 'Unsuperverised Learning' to identify the natural colour clusters within an unlabelled data set. The iterative process of the K-means algorithm grouped at least 5 colours in our test dataset. The source code can be found on my Github.

Skills & Tools

Software: MATLAB

Collaborators

Akansha Nagar, Tyler Nagata