Janakiram Sundaraneedi https://github.com/ram***

Mobile: 978-***-**** http://www.cs.uml.edu/~jsundara E Mail: ac4ss4@r.postjobfree.com https://www.linkedin.com/in/ram978 Available dates:Immediately

Visa Status: Post completion OPT (F 1 Status)

Education 2011 GITAM Visakhapatnam, AP, India

Bachelor of Science

Major: Information Technology

2017 University of Massachusetts, Lowell, MA,USA

Master of Science, Major:Computer Science (Graduated Fall 2017)

Graduate Course Work Advanced Algorithms, Foundation of Computer Science, Internet Web systems, Computer networks & security, Machine Learning, Xinu Operating System II, Data Mining, Data Base l, Heterogeneous Data Visual Analytics, Computer Graphics I, Big Data Design(Deep Learning), Human Computer Interaction and High Performance Computing(GPU’s).

Skills Programming: R, Python, C, MATLAB Databases: SQL OS: Linux, Windows, MacOS Other: Microsoft Office. Machine Learning: Regression’s, Decision tree, Random forest, SVM, PCA, KNN, Clustering’s, Boosting, Neural Networks. Deep Learning: Auto encoders, Restricted Boltzmann Machines, Deep Belief Networks and LSTM Network. Python Libraries: NumPy, Pandas, Scikit, Keras Web: HTML5, CSS, JavaScript Build Tools: Jupyter, Anaconda,TensorRT.

Experience Teaching Assistant for Analysis of Algorithm April 2015 – June 2015 University of Massachusetts, Lowell, MA, USA

Director Of Data Science(Volunteer) July 2017 – January 2018 Machine learning Society Boston Chapter

Set up meet ups with data scientists and discuss on various applications.

Decode latest products which are build using machine learning and Deep learning algorithms.

Work on research papers from IEEE and case studies from Nvidia labs.

Compete in Kaggle Competitions and Hackathon’s.

Decode mathematical equation’s for machine learning and Deep learning algorithms.

Work on new ideas and its implementation’s.

Project’s Type II Diabetes Prediction

• The ultimate goal to predict diabetes using web based application as front end and machine learning algorithms as back end.

• Using random forest classifier accuracy 76%, log loss 0.38.

• Using FCN accuracy 76%, log loss 0.42 and 80 epochs. Handwriting Recognition in Convolutional Neural Networks

• The objective is to distinguish hand written characters by running through training samples applying Stochastic Gradient Descent algorithm, Deep learning and Convolutional Neural Networks.

• Keras Deep Learning Python library was used on MNIST dataset and achieved 98.2% accuracy. Sentiment Analysis

• Develop a Support Vector model to classify the sentiment of movie reviews as either positive or negative. Stock Market Prediction based on Linear Regression

• Stock prices prediction was done based on the basic concepts of Linear regression on Gradient Descent algorithm based on the S&P 500 dataset.