Shujuan Ji

. AI Design: Developed blueprints for several AI products, analyzed the strategy of each development process, evaluated the pros and cons of applying different AI algorithms, and conducted cost estimation. One of the projects was recognized as an "Exemplary Assignment" of the MITxPRO class "Designing and Building AI roducts and Services ”.

DL, ML, and Data Science Projects in Python:

Implemented Gradiant Descent, CNN, RNN, DNN,LSTM, Transformer both by using plain Numpy, and by TensorFlow, Keras packages. Optimize the algorithms by proper initialization, regularization, minibatches, etc. Applied the algorithms to various applications such as a DNN classifier, face recognition, Neural Style Transfer. Adapted a pre-trained to new data.

Modeled, formed hypotheses, and performed statistical analysis on real data.

Used dimension reduction techniques such as principal component analysis, t-distributed, stochastic neighbor embedding to visualize high-dimensional data and applied this to genomics data.

Analyzed networks and used centrality measures to describe the importance of nodes, and applied this to criminal networks.

Modeled time series using moving average, autoregressive, and other stationary models for forecasting with financial data.

Used Gaussian processes to model environmental data and made predictions.

Communicated analysis results effectively with written reports and graphs created by Python.

Implemented a classifier to use for sentiment analysis of product reviews, compared and analyzed various algorithms, conducted parameter tuning.

Implemented multiple-digit recognition solutions using Neural Networks and other technologies.

Built a mixture model for collaborative filtering.

Implemented Q-learning algorithms to learn control policies for a text-based game.

·       Prompt Engineering: Prompted mathematical and logical questions to train AI models in step-by-step reasoning.

·       Web Development: Created web applications using Python, Flask, PostgreSQL, and JavaScript.

·       Math Tutor: Tutored grade 7-12 students on math enhancement, AMC preparation, and Calculus classes.

·       Skills: Python, Flask, Pandas, Pytorch, matplotlib, scipy, h5py, sklearn, tensorflow, keras, yad2k,JavaScript, PostgreSQL, Linear Classifier, SVC, Nonlinear Classifier, Kernel Methods, Feature Engineering, PCA, Neural Network, RNN, LSTM, CNN, Gaussian Mixtures, MLE, K-Means, K-medoids, EM algorithm, Reinforcement Learning, MDP, Hypothesis Testing, Unsupervised Learning, Clustering, Graph Centrality, Time Series, Seasonality, Stationarity.