Jayavarshini Ilarajan

Linkedin Github

About Me

I am an Analytics Enthusiast with around a year of professional work experience. As an Aspiring Data Scientist I am geting in-depth knowledge in Machine Learning and Visualization Techniques.

Academic Backgroud

M.S in Computer Science (Illinois Institute of Technology),Expected May 2019
B.Tech in Information Technology(Anna University) May 2017

Professional Background

Data Analyst (Aptar, Greater Chicago Area) Aug'18-Dec'18
Data Analyst (CodeBind Technologies, India) Nov'16-Jan'17

During this experience I have been exposed to the power of Analytics and learned how analytics can help shape the future of the organization. As a budding Data Scientist, I am learning about various Machine learning Techniques.

Other Technical Skills:

Languages: Python, SAS, SQL, Java, PHP, HTML, CSS,MATLAB
Tools:RapidMiner, SAS(Studio), Tableau, PowerBI, Git, MS Excel(PIVOT, VLOOKUP)
Databases: AWS Amazon Redshift, MySQL, Microsoft SQL Server Analytical/ Statistics
Skills: Regression, Clustering, Hypothesis Testing, Regularization, ANOVA, A/B Testing

Selected Independent Projects

Demand forecasting on Andes to use descriptive statistics for Drifter. Used demand distribution data by comparing the actual demand vs the forecast. Measured the forecast performance using the A/F ratio. Reduced the cost of Demand Uncertainty through Accurate Response to early sales.Language: Python| Computational Environment: IPython notebook| API and Libraries: NumPy, SciPy, scikit-learn
Sentiment Analysis of Amazon Echo Dot : Extracted features from 30,000 customer reviews from web scraping; Preprocessing vectorization in R using Text2vec. Tokenized using NLP; Compared performance of Deep belief Network and used SVM for Prediction.
Google Merchandise Store Revenue Prediction using Gradient Boosting, XG Boosting and Random Forest: Developed a predictive Model to estimate the future revenue of each customer in Google Merchandise for the Month of December 2018 and Jan 2019. Implemented a stacking model using Light Gradient Boosting and XG Boosting.
Neural Networks: CIFAR-10 data set Object Detection: Built a deep learning neural network which classified color images by optimizing hyperparameters using Bayesian Optimization and performed 20% efficient that the conventional convolution network. Tools and Languages: Python,PyTorch, TensorFlow