The goal of our study is to develop "BurgerBot," a framework combining GPT-4 and a segmentation model to guide a robotic arm in assembling a plastic burger toy. The objective is to understand real-time human-robot conversational interactions, focusing on the robot's adaptability to diverse instructions.

The goal is to build an end-to-end NLP system involving collecting our own data and training a model to identify specific entities such as method names, task names, dataset names, metric names and their values, hyperparameter names and their values within scientific publications from recent NLP conferences (ACL, EMNLP, and NAACL).

Develop a minimalist version of BERT (Bidirectional Encoder Representations from Transformers), implementing some important components of the BERT model (self attention, layers, model, optimizer, and classifier) to perform sentence classification on sst dataset and cfimdb dataset.

In this course, I worked with Philips in a group of students to help them become more entrepreneurial. As part of the corporate strategy team, we evaluated new business ideas to boost their market share and revenue. This involved analyzing customer personas, market size, MAP planning, and value innovation, as well as developing skills for pitching to the C-suite.

The focus of this group project is to implement, evaluate, operate and monitor a movie streaming recommendation service in production, which entails many concerns, including deployment, scaling, reliability, drift and feedback loops. The streaming service has about 1 million customers and 27,000 movies.

In this Kaggle competition, I learned how to build an encoder to effectively extract features from a speech signal, how to construct a decoder to sequentially spell out the transcription of the audio, and how to implement an attention mechanism between the decoder and the encoder.

In this Kaggle competition, the task was to build a face classifier that can extract feature vectors from face images and a face verification system that computes the similarity between feature vectors of images. I used a CNN architecture to build this model in order to achieve high accuracy on this classification and verification task.

In this Kaggle competition, the task was to predict the phoneme label for each frame in the test set of the speech recordings, which are raw mel spectrogram frames. I used a multi-layer perceptron model and explored various hyperparameters to improve the accuracy of the prediction of the phoneme state labels for each frame in the test set.

I implemented an RNN-based language model, text prediction and degeneration, greedy decoding, and regularization techniques for RNNs. Here, I go over these concepts and a demonstration of language modeling for machine translation (English to French).

I implemented RNNs, GRUs, and CTC from scratch. Here, I go over a binary sentiment prediction task using the TensorFlow IMDB dataset, give an overview of these architectures and compare RNNs, GRUs, and LSTMs.

I implemented convolutional neural networks (CNN) from scratch. Here, I go over a traffic sign identification task and describe the significance of convolutional layers, pooling layers, downsampling and upsampling layers, classification layer, forward pass and backpropagation through layers.

I implemented an MLP from scratch with 0, 1, and 4 hidden layers. Here, I go over an MLP sentiment analysis task and describe the significance of linear layers, activation functions, forward inference, criterion functions (MSE and CELoss), SGD optimization, batch normalization regularization, and backpropagation.

This course is both a survey of computer law and an examination of how courts and administrative agencies make decisions on issues involving computer technology. The material is divided into these six primary subjects: legal process, evidence, eBusiness law, personal intrusions, intellectual property, government regulation. Here, I write key insights from these six topics and my personal reflection.

This course aims to teach students about AI techniques across various applications, generate new product ideas in specific fields, and review AI company trajectories. In a team of 4-5, I researched and designed a presentation on the use of AI in randomly assigned fields such as Identity Verification, Data Privacy, Firefighting, Personalized Medicine, Perfume, and Advertising.