Chaitanya Madhu
ML Engineer Data Scientist Innovation Lead
966-***-**** . ada0g3@r.postjobfree.com
Automated the pipeline flow of two major Startups using Data Science innovations. Looking to work with a fluid nerd team and recreate new ML strategies. Skills
Machine Learning
Python - SVM,Decision
Tree, Random Forest,
KNN, Logistic Regression,
xgboost, K-means,
DBSCAN, Hierarchical
Clustering.
NLP
Python - NLTK, Spacy,
TextBlob, Vader
Sentiment, Word2Vec,
Seq2Seq.
Data Science
Python - Pandas, Numpy,
Scipy, Matplotlib,
Scikit-learn, OS, Seaborn,
Dlib, Open CV.
Aldsgdlsgka;
lkga;gjhjalsdflgfalg;k
Data Generation
Generative Adversarial
Networks (GAN), DCGAN,
LDA, t-SNE.
Deep Learning
Python - Tensorflow,
Keras. Multi-layer
Perceptron, CNN, RNN,
LSTM, GRU, Seq2Seq
models and modern
activation functions.
Databases
Sql - My Sql
NoSql - Mongo DB
Experience
Bricolage Technologies— NLP & Text Mining Internship August 2019 - November 2019
● Contributed to Automate the entire pipeline processes of this US-based Fashion Trends Forecasting Startup using NLP & NLU by means of a custom built library - Bric
(Enhanced over NLTK, Spacy and Textblob).
● Framed the custom library Bric in a flexible and extendable way to meet the future requirements of the company.
2
● Helped the company to cut-off its overdue expenditure (around INR 2,00,000 per month) for Crawling and NER tasks with the indulgent of Machine Learning algorithms.
C-Affine AI — Machine Learning Internship (NLP & CV) September 2018 - July 2019
● Introduced new algorithmic approaches for the mood-sharing chatbot project -
"Advanced Digital Personal Assistant - Caffine".
● Implemented LSTM,GAN and CV techniques to build NLU and NLG sections of the digital personal assistant.
● Supported the project for a long period(about 1 year) in creating an effective human-computer interactive design.
Projects
Chatbot using Seq2Seq—
● Developed an advanced chatbot with training data from Cornell movie dialog corpus.
● Used Tensorlayer along with Tensorflow to get the task done with less time-complexity.
● Also invoked new activation function aggregations to generate effective replies in no time. https://drive.google.com/open?id=1yow0wJsANZ1B7QQ845wdJJGuJ16nga1i
Data Generation using GAN & DCGAN—
● Designed the whole project under two sections:
a) GAN for Music Generation b) DCGAN for Human Faces Generation.
● Used GAN on Piano (mono-instrumental) tunes to generate new tunes of the same style.
● Applied DCGAN on human faces dataset to generate fresh artificial human faces.
● Improvised the existing GAN algorithm, eliminating repeated sequence operations which resulted in reduced time-complexities and boosted performance. https://drive.google.com/open?id=1bTG0SKapQjVn5Io5BOWaziMpc01Kp4pJ 3
Stock Price prediction using LSTM—
● Implemented LSTM network to predict the future stocks of Apple Inc.
● Also tried implementing a new version of the Stock price prediction model by considering the aggregate of both previous stock trends dataset and company business vulnerabilities dataset.
● Implemented a sequel to the project but using the time-series approach for a much efficient and precise value predictions.
https://www.kaggle.com/chaitanyacc4/predicting-stock-prices-of-apple-inc/data
D Drowzy using Open CV—
● Modelled techniques using OpenCV for the tasks of face & eye detection.
● Used Euclidean distance measure to keep track of the degree of eye-lid movements and dizziness accommodated with substantial time intervals.
● Also associated the project with an android application to alert drowsy vehicle drivers to prevent road crashes.
https://drive.google.com/open?id=1KwupIGGgfWPEAlBm4_9j59SySo8V9-yL Education
Degree Name of the Institute CGPA
B Tech - Computer Science &
Engineering
(year of pass: 2020)
IIIT-Nuzvid 9.2/10 [2019]
Senior Secondary (Pre-University
Coaching)
IIIT-Nuzvid 9.0/10 [2016]
Secondary (High School) SPPPRZPHS-EM
(Pandalapaka)
9.8/10 [2014]
4
Awards
● Winner of National Level Paper Presentation contests(on Machine Learning Enhancements) held at GEC & RGUKTN [2018]
● Best Speaker Award Recipient at Techzite'18 held at IIITN (in a combat of 500 participants) [2018]
● Prathiba Award Recipient [2014]
Languages
Python, C, Java and proficient in English, Hindi and Telugu.