Software Engineer

AKASH DEO

**** ********* ****, **********, **-***80 469-***-**** ******.*****@*****.*** www.linkedin.com/in/akashdeo EDUCATION

University of Texas at Dallas, Richardson, TX, USA (August 2016 – May 2018) Master of Science, Computer Science, May 2018 [3.3 GPA] Shri Ramdeobaba College of Engineering and Management, Nagpur, India (July 2012 – May 2016) Bachelor of Engineering, Computer Science, May 2016 [8/10 CGPA] SKILLS

Programming Languages: Java, C++, C, MySQL, Scala, C#, R Scripting Languages: Python, bash, PowerShell

Web Programming Languages: PHP, JavaScript, jQuery, AJAX, HTML5, CSS3, JSON, XML, Node.js Natural Language Processing Tools: SOLR, Apache Open NLP, Stanford Core NLP. Big Data Technologies: Hadoop(Map-Reduce), Apache Spark-SQL, Twitter data streaming. Security Concepts: TCP, UDP, Cryptography, Authentication. Other Frameworks/Libraries used: Maven, Git, Android SDK, Face SDK, Junit, MLLib. Operating Systems worked with: Windows, Linux, OS-X WORK EXPERIENCE

Teaching Assistant (Erik Johnson School of Computer Science, University of Texas at Dallas)- (Aug 2017- May 2018)

• Responsibilities: Grading assignments and helping undergraduate CS students understand concepts like variable assignments, loops, file handling, user defined functions, arrays, objects and classes, pointers in C++. PROJECTS (https://github.com/akashdeo/)

Semantic search application that will produce improved results using NLP techniques (UTD - Fall 2017) The objective of this project was to create a sophisticated search engine that would yield high quality search results by combining NLP features in addition to primitive keyword search. Three tasks were done:

• Simple search pipeline that used only keywords. ( https://github.com/akashdeo/NLP-Fall-17-Project )

• The search index was made using features like hyponyms, hypernyms, part of speech tags etc. using Stanford CoreNLP Package instead of just using keywords. Index was built using SOLR tool which is based on JAVA. Language used: JAVA Camping Reservation System ( https://github.com/akashdeo/CampingReservation ) (UTD - Fall 2017)

• Website allows the user to register / login into the reservation system and then select the camp whichever available on the website. Cart and Wishlist functionalities were implemented.

• Users can cancel camps at any time as well. Admin can add as well as delete camps at any time and proper message is sent to the users regarding camp cancellation.

• Technologies used: JavaScript, AJAX, PHP, jQuery, HTML5, CSS, Bootstrap. Software Metrics Calculation of any Java GitHub Project. (UTD - Spring 2017) (https://github.com/akashdeo/SoftwareMetrics )

• It is an Apache Maven plugin used to measure the code metrics like Number of Operands, Variable References, Class References, Method Name, Number of arguments. (Language used: JAVA) Calculation of Accuracy of Letter Recognition Dataset using Different Machine Learning Algorithms. (UTD-Fall 2016)

• The objective is to identify each of many black-and-white rectangular pixels displays as one of the 26 capital letters in the English alphabet. We typically train on the first 16000 items and then use the resulting model to predict the letter category for the remaining 4000.

• Took a suitable dataset from UC Irvine Machine Learning Repository and ran different Algorithms like Support Vector Machine, Random Forest, Decision Tree, K means, Bagging, boosting etc. to calculate the accuracy. The Maximum Accuracy was obtained by Support Vector Machine (>90%) followed by Bagging and Random Forest. (85-90%). Decision Tree also gave around 80% accuracy. (Language used: R) Pacman Projects from CS 188 course in UC Berkeley (UTD – Fall 2016) (https://github.com/akashdeo/AI_Pacman_Projects ) These projects were done as a part of Artificial Intelligence Class in UT Dallas. (Language used: PYTHON)

• Different search algorithms like Breadth First Search, Depth First Search, Cost Based Search were implemented in the first project. These algorithms helped calculate the cost of pacman eating all the available food.

• Second project comprised of Min – Max algorithm and the third project was based on probabilistic techniques like eating the food according to where the ghosts are located and maximizing the cost in that context. Implementation of NLP/Machine Learning Algorithms (UTD-Fall 2017) (Language used: JAVA)

• Implemented Brill’s Tagger Algorithm which tags proper part of speech to all the words in a sentence. 100% accuracy achieved. ( https://github.com/akashdeo/Transformation_based_POS-Tagger )

• Calculation of N-gram Probabilities in each sentence using Smoothing and without Smoothing.

• Implemented Lesk Algorithm to identify the correct part of speech tag of a word in a sentence using Wordnet.

Contact this candidate