ML, AI & Data analysis related Projects
PARTIALLY OBSERVABLE MARKOV DECISION PROCESS ALGORITHM
Predictive model to predict position of Robot in a grid.
Developed a POMDP algorithm that can predict the position of a robot in the given grid using JAVA
Independently developed the observation model and the transition model required to calculate the probability of being in a certain state (belief state) and update it after every action and taking an observation.
Bank Customer retention analytics
Python and reinforcement learning
Prediction model on the likelihood of a customer leaving the bank. Developed a DNN using kerasAPI, tensorflow, and numpy libraries in Python, training the DNN model on a given data set and calculating the probability based on multiple selected factors.
BILL-SPLITTING APPLICATION
Using JAVA OOP and JSON implementation to create an item specific bill splitting app.
Using self-developed algorithms, created an interface which allowed users to split the bill accurately, instead of just dividing the total amount equally among the group.
Each item in a bill is split with respect to only the people that that used the specific item.
