syedhadi816

Automated-Essay-Scoring_BERT_CNN_BiLSTM

Second-Order-Central-Finite-Difference-

The attached code can be used to solve a second order ordinary differential equation using the central finite difference method. It has currently been written to solve the second order light deflection ODE.

ARIMA-and-Experimental-KNN-for-time-series-data

An ARIMA time series forecasting model to forecast the LIBOR rate during the COVID pandemic. For experimental and learning purposes the Feds Funds rate was forecasted using ARIMA and the LIBOR forecast was calculated using the FFR forecast through KNN (which surprisingly gave good results). But that's mainly because FFR fluctuations are very closely tied with LIBOR, this approach may not be used for conventional time series forecasting.

BackProp-Pytorch-Hand-Written-Number-Recognition

MNIST Numbers dataset to train a model on Pytorch to detect hand written numbers

Graph-Convolutional-Network-for-Bundle-Recomendation

The model described in the paper was trained on the goodreads books dataset to imitate the recommendation of bundles of online courses to users.

Modeling-the-Compressive-Strength-of-Concrete-with-MARS-and-Ensemble-RF

Implementation of a statistical model for measuring the compressive strength of concrete which explains 92.3% of the variance.

Monte-Carlo-Simulation-for-Black-Scholes

Monte carlo simulation for the black scholes PDE.

Sarcasm-Detector-

Sarcasm detector for news headlines using KNN, Log Regression and Perceptron Neural Network

Binomial-Tree-for-Options-Pricing

Binomial tree for European put option.

Cox-Ingersoll-Ross-CIR-Simulations-

The following code can be used to obtain a Cox-Ingersoll-Ross (CIR) simulations. I created the code to obtain a simulation for interest rates and house prices that were to be used to numerically solve a mortgage pricing PDE.

derby-race-prediction-

Horse race analysis and modeling to understand factors that affect the final position of a horse in a race. This includes analysis of track conditions, jockey (and jockey history), horse history, course type and a number of other factors.

digital-psychotherapist-chatbot-on-Python-with-Regex-from-Scratch

human-activity-recognition-with-deep-learning-multi-class-time-series-classification

This repository is currently a work in progress.

K-Nearest-Neighbours--No-Libraries-Used-

KNN (Spring 2020)

Logistic-Regression-No-Libraries-Used-

Logistic regression with softmax activation and mini batch gradient descent. (Python)

Model-for-Fair-Auction-Price-of-IPL-Players

An ML model for determining the fair auction price of cricket players for the IPL. If you wish to replicate this model reach out at syedhadi816@gmail.com for data files used. Data cleaning and preparation was the most intensive time intensive component of this project.

Monte-Carlo-Integration-

Estimating the solution of an integral using monte carlo integration.

Naive-Bayes-Classifier--No-Libraries-Used-

pipedream

Serverless integration and compute platform. Free for developers.

Retreival--Generative-Ensemble-Conversational-AI--BERT-Transformer-and-BiSeq2Seq-w-Attention-Layer

RNN_short_text_generation

Character Level Language Model

sequence-tagging-NER-with-feature-generation-parameter-optimization

Similarity-Measurement-in-Text-Data

Measuring similarity in textual data (Bag of Words model) using Jaccard distance, Cosine similarity and Euclidean distance.

Stochastic-Gradient-Descent--No-Libraries-Used-

Logistic regression using log loss and stochastic gradient regression from scratch.

SVC-with-varying-kernels-for-wine-classification

syedhadi816

Config files for my GitHub profile.

text-classification-with-ensemble-RF-and-RandomSearchCV

UNet_Image_Segmentation

UNet Model for Image Segmentation.

Wikipedia-Answer-Generation-Regex-Pattern-Matching

This QA system should be able to answer Who, What, When and Where questions (but not Why or How questions) and provide the information only what is asked for by the question. below is the sample question and answers for that.