Thanushree Eadala Gajendra

Highly-motivated employee with desire to take on new challenges. Strong worth ethic, adaptability and exceptional interpersonal skills. Adept at working effectively unsupervised and quickly mastering new skills.

• Microsoft Azure Fundamentals

• Awarded as the Best Collegiate Social Responsible Leader for the tenure of my bachelors for the social service activities in and around the university for leading events like Swachh Bharat, Plantation Drive, Blood Donation Camp, Flood Relief Drive, Visit to Orphanages and many such things which happen in our campus.

Lung Cancer Detection Using Gxnor_Net:(Final Year Project)

• Description: A framework for Lung Cancer Nodule Detection Using GXNOR_NET. Dataset called LUNA-16 related to Lung Cancer is used. In order to minimize the cost function using Gated XNOR[01] which discretizes the weights as well as use and activations of back propagation with ternary neural activations on the above mentioned dataset as well as in order to view the behaviour of the networking by altering some of the network parameters.Technologies: Python2.7, Flask,

SelfiLessActs App.

• Description: A cloud based web application built on AWS EC2 instance using container orchestration that can perform load balancing, fault tolerance, and auto-scaling.
• Integrated the cloud backend with the mobile application. The app that is used to share information about anything that is good for the society. The application will allow users of the application to upload image of the act with a small caption and a categories.
• Allows them to select a topic. Upvote a particular Act. Upload an Act. Delete an Act. Technology: Amazon EC2 instances, Flask web framework API: REST

Deep LSTM Siamese Network For Text Similarity:

• Description: A tensorflow-based deep siamese LSTM network that uses character embeddings to detect related phrases and sentences. Code provides a framework for learning phrase and sentence similarity using character- and word-level embeddings.
• A multilayer Siamese LSTM network and Euclidian distance-based contrastive loss determined how similar two input pairs were.
• This model can learn structure and semantic similarity with enough training pairings. "He is smart" indicates "He is a wise man." I.B.M. stands for International Business Machines. To detect commonalities, character-based and learnt word embeddings are used.
• Phrases training took 6 minutes, 48 seconds, and was 91% accurate. Training sentences took 8 minutes and 10 seconds and was 81% accurate. Technologies: Python

Hospital Management System:

• Description: It was a project developed to maintain the database of the hospital systems. This project consists of all the aspects an hospital requires to maintain its data and store the data in the back end. Technology: MySQL, Html, PHP, JavaScript.

Online Food Recommendation System:

• Description: It was a food recommendation system using K-means Clustering algorithm for the recommendations. Data set was taken from Yummly website. The website is under construction yet. Data is in JSON format.
• On our get recommendation function, once we feed the food name as the input, they are tf-idf vectorised. These vectors are sent as inputs for k- means clustering algorithm for prediction of similar foods on the list.
• The predictions are then stored in our data frames for further display. Technologies: Flask, HTML, CSS, Ajax, Jinja, jQuery, Bootstrap.