Certificate in Quantitative Finance (CQF)
Here are some financial quantification projects I did while studying CQF:
- three_asset_compare: Google stocks(GOOGL), Gold, and Oil are used here for comparison. The conclusion is that Gold has a positive correlation with Oil and has a negative correlation with The stock of Google.
- FP_groupA: Use LSTM model to train and predict GOOGL
- FP_groupB: Use LSTM model to train and predict Gold
- exam3_Group_A: Group A Stock only predicts the direction of very short-term returns; Group A is a binomial classification, moves close to zero will show up in daily returns. If we bundle them with negative values, this will most likely produce overpredictions of negative returns, so you can choose based on your data.
- exam3_Group_B: Group B time series is an index, factor, interest rate, or economic quantity for which you will be predicting 1M returns or changes (the longer-term) – one ticker or quantity. Given that the logistic classifier is formulated for binomial/multinomial prediction, it is recommended to explore a histogram and set up multinomial scheme, eg to predict buckets not necessarily a numerical value.
- demo1: Plot Bernoulli distribution, binomial distribution, geometric distribution, Poisson distribution, normal distribution, Monte Carlo simulation
- Assigment_#1:
- Design two different strategies() to carry out the Monty Hall decision problem. Compare the results.
- Prove the Central Limit Theorem through a simple experiment. Compare experiment with the theoretical equivalent gaussian curve.
- Assigment_#2: use K-Nearest Neighbors pretain and test random generate dataset, calculate and compare Generalization Error (GE), Model Prediction Error (ME), Training Error (TE).
- exam3:
- Section 1 Classifier and Hyperparameters
- Section 2 Model Selection
- Section 3 Mathematical bases of supervised learning