Neural Networks and Deep learning Project

Semantic Segmentation of Thermal Images using U-Net algorithm (To be revised)

(To be revised) Achieved a binary cross-entropy loss of 0.118 against a benchmark of 0.0172, for segmentation of IR images collected from the university’s ADAS dataset trained on the UNet algorithm using Jupyter Notebook

Data Mining Project

Binary Classification model using Logistic regression in R

Achieved 77.5% accuracy in the binary classification of fire incidents using Logistic regression trained on the Boston fire dataset

Project Management

What does an organization need to bring JARVIS into conception

Draf of project organization and implementation plan (IDEF0 Model, WBS, Gantt chart, PERT, RACI, and stakeholder matrix) for an AI/ML backed smart home assistant product development at a target budget of $2.0 Million