Control Traffic Data Scientist

Kanimozhi Kanagaraj

Livermore, CA ********@****.***.*** linkedin.com/in/kani-sharath/ 231-***-**** Technical Skills

Programming: Python Java C++ Html Mysql Javascript Vhdl Django Data Analysis Machine-Learning Algorithms: Linear Regression Svm Algorithm Naive Bayes Algorithm KNN Algorithm K-Means Clustering Convolutional Neural Networks Q-Learning

Machine-Learning Frameworks: Tensorflow Pytorch Keras Sci-Kit Learn OpenCV Soft Skills: Teamwork Problem Solving Adaptability Flexibility Communication Self-Learning Education

Data Scientist Certification - Fellowship Program, The Data Incubator Expected: Nov 2022 M.S. in Computer Science Loyola Marymount University GPA: 3.93/4.0 May 2022 M.B.A. Finance & Marketing Concentration Dr. G.R. Damodaran College of Science GPA: 3.39/4.0 May 2018 B.E. Information Technology Avinashilingam Deemed University GPA: 3.17/4.0 May 2016 Experience

CALQLOGIC Sep 2022 – Present

Data Analytics & Machine Learning Intern

• Defined report on API for Healthcare data extraction, targeting accurate results, efficient operation, and low cost.

• Developed python scripts for web scraping using API’s to access large datasets.

• Developed python scripts to do sentiment analysis on text formats using API’s

• Secured contracts by educating clients and partners on software, methodology, and superior results versus competitors. HDFC HOME LOANS Dec 2017 – Feb 2018

Technical Consultant Intern

• Assisted team to approve customer loans. Recommended company app to update agents, based on statistical analysis.

• Conducted a survey of third-party agents to understand their understanding of products and partnership satisfaction. METTLER TOLEDO TURING SOFTWARE May 2014 – Jul 2014 Software Intern

• Developed code in Java for an X-Ray inspection system used for examining packaged products.

• Collaborated with the intern team to debug java code. Trained in basic unit testing at an industrial level. Projects

Capstone: Traffic Sign Detection & Classification, Deep Convolutional Neural Networks Spring 2022

- Classified traffic signs by applying deep learning CNN model with VGG net architecture in Python.

- Achieved 95% accuracy on new web images with the model. Utilized German Traffic Sign Benchmark as image data.

- Leveraged TensorFlow backend with OpenCV to manipulate images.

- Accomplished training accuracy of 98% and achieved testing accuracy of 98%.

- Addressed overfitting with image augmentation and early stopping. Analyzed performance via SoftMax probabilities. Traffic Signal Control Using Reinforcement Learning – Q-Learning Algorithm Spring 2021

- Developed a Python code to control traffic light phase using reinforcement learning via a learning agent.

- Implemented a reward system that led to agents controlling the traffic lights efficiently, thus optimizing traffic flow.

- Achieved a successful computer simulation with SUMO software. Designed Website “LIKEYOU” Spring 2021

- Developed a website that recommends influencers based on customers skin tone

- Website successfully guided users to relevant YouTube influencers

- Agile methodologies were used to manage project as a team

- Accomplished effective teamwork using scrum methodology Image Segmentation Using Fully Convolutional Neural Networks Fall 2020

- Developed Python code to train a CNN that produces per-pixel category labels of an image, using the cityscape dataset.

- Produced correct image partitions with the model, with accurate categorization. Evaluated with Jaccard Index and intersection over union. Sentiment Analysis on Movie Data Fall 2020

- Sentiment Analysis of movie reviews using Stochastic Gradient Descent (SGD) classifier

- Successfully categorized movie reviews as positive or negative using reward systems

- Random_search_cv method was successful in finding the best parameters and scores

Contact this candidate