Pinned Repositories
CE807-Text-Analytics-Toxicity-Classification-
This project compares generative and discriminative models for toxicity classification in text comments. Use Google Colab to upload your dataset and run the provided scripts to train and evaluate both models. For setup, install dependencies with !pip install transformers and follow the script instructions.
CE811-lab
CE888-Data-science-decision-making
CE889-Neural-Network-for-Rocket-Landing-Game-No-External-Libraries-
--- This project implements a feed-forward neural network with backpropagation from scratch, without external libraries. It trains the network to control a rocket's landing in a simulated game using data collected from manual gameplay. The trained network is then used to guide the rocket's landing in real-time.
CE889-Rossmann-store-sales-Group-project-
This project uses Long Short-Term Memory (LSTM) networks to predict Rossmann store sales. Implemented in `Neural_Networks_GroupAssignment_LSTM.ipynb`, it covers data preprocessing, LSTM model training, and evaluation. The notebook includes steps for cleaning data, building the LSTM model, and assessing performance. ---
CE903-TEAM27-NB-IoT-LTE-M-Systems
This project enhances indoor positioning by combining Observed Time Difference of Arrival (OTDOA) with NB-IoT and LTE-M technologies. It addresses GPS limitations indoors, offering precise and efficient location tracking for objects and individuals within buildings.
ECNN-Enhanced-convolutional-neural-network-for-Automatic-ECG-based-Classification-
Handwritten-Digit-Recognition-Training-and-Evaluating-a-Neural-Network-numpy-no-TF-Keras-
Here’s a concise description under 300 characters: --- This repo features a neural network for handwritten digit recognition built from scratch with `numpy`, without TensorFlow/Keras. It includes data preprocessing, model training, and evaluation, achieving ~89.9% accuracy on the MNIST dataset. Ideal for learning core neural network concepts. --
Navedshk's Repositories
Navedshk/CE807-Text-Analytics-Toxicity-Classification-
This project compares generative and discriminative models for toxicity classification in text comments. Use Google Colab to upload your dataset and run the provided scripts to train and evaluate both models. For setup, install dependencies with !pip install transformers and follow the script instructions.
Navedshk/CE811-lab
Navedshk/CE888-Data-science-decision-making
Navedshk/CE889-Neural-Network-for-Rocket-Landing-Game-No-External-Libraries-
--- This project implements a feed-forward neural network with backpropagation from scratch, without external libraries. It trains the network to control a rocket's landing in a simulated game using data collected from manual gameplay. The trained network is then used to guide the rocket's landing in real-time.
Navedshk/CE889-Rossmann-store-sales-Group-project-
This project uses Long Short-Term Memory (LSTM) networks to predict Rossmann store sales. Implemented in `Neural_Networks_GroupAssignment_LSTM.ipynb`, it covers data preprocessing, LSTM model training, and evaluation. The notebook includes steps for cleaning data, building the LSTM model, and assessing performance. ---
Navedshk/CE903-TEAM27-NB-IoT-LTE-M-Systems
This project enhances indoor positioning by combining Observed Time Difference of Arrival (OTDOA) with NB-IoT and LTE-M technologies. It addresses GPS limitations indoors, offering precise and efficient location tracking for objects and individuals within buildings.
Navedshk/ECNN-Enhanced-convolutional-neural-network-for-Automatic-ECG-based-Classification-
Navedshk/Handwritten-Digit-Recognition-Training-and-Evaluating-a-Neural-Network-numpy-no-TF-Keras-
Here’s a concise description under 300 characters: --- This repo features a neural network for handwritten digit recognition built from scratch with `numpy`, without TensorFlow/Keras. It includes data preprocessing, model training, and evaluation, achieving ~89.9% accuracy on the MNIST dataset. Ideal for learning core neural network concepts. --