NLP Module

For managers in companies who have a complex manual process in need of automation. I am an algorithm scientist with a Masters Degree in Signal Processing from Purdue University, and a penchant for outside-the-box thinking. I offer innovative algorithms designed to solve challenging and costly problems. I have eleven years experience designing algorithms, and pull from a rich background of skills and knowledge including: Machine Learning, Computer Vision, Image Processing, Genetic Algorithms, Neural Networks, Regression, System Design and Modeling, Predictive Analytics, Python, OpenCV, TensorFlow, and Deep Learning. Unlike other algorithm scientists, I have leadership and communication skills that can be leveraged to help guide teams of developers to implement customer-facing or real-time versions of the solutions provided.

Hello, I have good experience in handling text data. I have completed various projects of handling text data like spam-ham classification, recommendation, review positive/negative classification etc. I can help you design a DL/ML model which classifies intent of utterance. Please feel free to go through my profile to see my past experience and skill-set. Please feel free to contact me to discuss further about it. Regards, Sunny

Hi, I am interested in your project and want to know the detailed info such as development environment, target language etc. I have been an export in this field for 5+ years, and have full experience. Waiting for your positive response ! Best Regards, Jin

*****Hello, dear!!!***** I have read your description carefully. I can handle it with full confidence and have already done this type of projects. Please give me an opportunity to work with you.

Hi I am Expert badge on kaggle, I have worked on a similar projects where my task was "intent mining " from chat complaints . Task-1 for the project was to classify the labeled intent. I used the Self attention model with Bi-direc LSTM and achieved an AUC of 83%. Task-2 for the project was get the key intent from each chat ( unsupervised problem ). In this case I used spacy + Nltk +word2vec model to get the intents from unlabeled dataset. We can discuss about bid price after the discussion of project work you can also refer to my kaggle profile (username: kanav0183) AV(username: kanav) /for more info.