Data Scientist at Precisely | Ex- Amazon, Zomato, Ericsson R&D, Baker Hughes | IIT Jammu | Mentor

github

Rajat Mittal  • Peerlist

Deployed IoT Sensors for Temperature & Humidity, Soil Moisture, Air Quality Index and Deflection with the main Gateway node.
Collected real-time data through sensors over a cloud using a small private network maintained in the gateway.
Forecasted weather (Temperature, Humidity) for the IIT Jammu campus using Machine Learning techniques.
Predicted landslides in the hilly areas using soil moisture and deflection sensor by setting some threshold value of deflection
in a particular axis and observing previous data of soil.

Tools: Python, Numpy, Pandas, Sklearn, IoT

Weather Forecasting System

Performed text preprocessing on web data over a corpus of 2000+ documents.
Preprocessed documents by removing stop words, tokenizing and all to get the processed document for further indexing.
Created a dictionary (key-value pair) consisting of DocID as key and the frequency of that word as the value in that document.
Created inverted index where implemented Boolean queries and Phrase queries to get the required documents.
Technologies Used: Python, Pandas, nltk

Data Organization and Information Retrieval System (IR Sytse

A Web Application using the Django Framework of Python. 
- Data fetching is done by RapidAPI
Web Application is live on:-
http://rajatenzyme.pythonanywhere.com/

Covid-19 Live Tracker of India

A live covid-19 Tracker website using HTML, CSS & Javascript. 

Covid-19 Tracker (Indian States and World Data)

->Developed a prediction tool of Notes and Instruments in a given Audio file in a team of four members.
->Used SVM classifier to detect the instrument being played in an audio file, along with all the notes and frequencies present in the audio file.
->Extended it to make a robotic system that would mimic the audio, based on the results given by the classifier.
->Reached in the pre-final round after clearing five competitive stages.
->Tools - Sklearn, NumPy, Pandas, Wave, Struct, ML Algorithms


Tasks accomplished:
Python:
    Take music file as input.
    Machine Learning to detect the instruments used.
    Use audio processing to detect note onsets and individual notes.
    Use serial communication to send this information to AtMega 2560.

Embedded C:
    Bot:
         Receive serial input from the Python script.
         Process input and control the bot to strike on piano and trumpet.
    Instrument:
         Capacitive touch sensor(placed on piano and trumpet) receives input on striking.
         Detect which note is being played.