Description
I have been working as a freelancer from past 4+ years and handled 20+ projects ranging from $100-$1000 of worth, covering Deep Learning, Machine Learning, Flask API, Cloud Deployment etc.- Primary Skills: Machine learning & Data Science
- Secondary Skills: Python, Pandas, SQL, Deep Learning, Cloud computing, API
- Built Machine Learning & deep neural networks.
- Used best practices to train-test sets & analyze bias variance for building DL applications with optimization-algorithms, end-to-end using TensorFlow.
- NLP-RNN Models Built Recurrent Neural Networks & its variants (GRUs, LSTMs), applied to character-level language modeling working onNLP with hyperparameter tuning strategies such as Dropout & Batch Norm etc.
- CNN ModelsBuilt a Convolutional Neural Network to use in visual detection and recognition tasks.
- Built a neural style transfer to generate art & applied these algorithms to image, video, and other2D/3D data.
- Object Detection Worked as ML intern to recreate a project like Amazon go, built a Computer Vision aided program to detect and classify objects in daily life.
- Data Science Deployed a Recommendation model to an international client from France. It was a machine learning model that maximizes hit rate from the data provided by him
- LSTM Model Financial (stock price prediction) Time Series Forecasting with LSTM. This was for an Australian client where I coded univariate and multivariate stock price prediction with LSTM in Python.
- Graph Plotting Data Analysis especially using Visualization &Graph analysis projects were handled. Libraries like Seaborn and Matplotlib for creating attractive graphs was used for around 3 consecutive projects.
- TF-Web-Apps Built Web-apps like a classifier which lets you draw in the browser & recognizes your handwritten digits and other was a toxicity classifier, which uses NLP to determine if a phrase is toxic or not.
- TensorFlow Lite Learnt Advanced Deployment scenarios with Tensor Flow including on mobile devices and IoT devices. Performed capstone projects and earned a specialization certificate.
- Created Flask APIs for a US client (CTO to be precise) & deployed it on his webserver hosting with Centos OS. Had a great experience working on Postman software to manage all of the APIs.
- AI Prediction API Same client further assigned me his next big project to make a recommendation price model for used car market using Artificially intelligent. Car’s specs are fed and API gets a price range for it.
- FRCNN Model Another US client approached me for an advanced computer vision project where the Faster-RCNN model was trained to detect diseases in leafs of different plant images.
- Intrusion Detection Guided handful of University students at Toronto for their project on subjects like Network Intrusion Detection system. Machine learning models were trained at Huge data with great accuracy results.
- CT-Scan Classify An interesting project was handed fora Canadian client where COVID-19 patients were classified by my CNN model. CT-scanned lung images were trained and tested for positive &negative lungs.
- GCP Big Data Learnt Google Cloud Platform DBMS. Migrated Cloud SQL & Data proc to existing MySQL & Spark/ Hadoop/Pig/Hive workload to Google Cloud. Employ BigQuery to do interactive data analysis.
- Business Use case Learnt How Google does Machine Learning, From Framing a business use case as a ML problem to gain a broad perspective of it & where it can be used. Recognized biases that a ML can amplify.
- Data Analyst a real business use case was accomplished by me. A British client had local food restaurant data with him. I did Exploratory data analysis of data science and gave critical insights & suggestions.
- AWS Basics Learnt AWS S3 basics. Features like Encryption, Versioning, Static Website. Worked with S3 Bucket and uploaded various forms of data Content, managing their Access and management rules.
- AI in Medical Many clients has approached me from different domains of expertise for doing their experimental statistical calculation and analysis required in their PhD thesis or paper publications.
- AI in Finance a couple of clients have approached me for Machine learning for stock prediction python code where historical data &live data are fed to learn the pattern and deduce future trends to investing.
