The widely-accepted issue, so-called: "The Exchange Rate Disconnect" or "The Exchange Rate Puzzle". This issue leads to great difficulties to predict this variable. Thus, in this project, I develop a Convolutional Neutral Network (CNN) algorithm. Accordingly, this simple AI project is used to predict for the Nominal Effective Exchange Rate (NEER) of the 12 EU-nations. The NEER series has the monthly period of January 1970 to September 2019. It is collected from the Bruegel Datasets. The trained AI model fits data well and makes a relatively accurate forecast. The Jupiter code of this project can be found my repository on the Github space below.
https://github.com/phuongvnguyen/The-AI-based-predictions-for-the-exchange-rate
Caution: This AI algorithm has not been optimized yet and still been under construction. Thus, one uses this Jupiter code for reference with caution.
In this project, I use the Sklearn datasets on breast cancer to train a supervised machine learning model. Accordingly, various machine learning model is used, such as Logistic Regression, AdaBoost and Gradient Boosting Classifier, Random Forest and ExtraTrees Classifier. However, the Linear Discriminant Analysis (LDA) algorithm delivers the highest score of prediction for if a patient suffers from breast cancer. The Jupiter code of this project can be found my repository on the Github space below.
https://github.com/phuongvnguyen/The-Breast-Cancer-Diagnostic
In this project, I train several supervised Machine Learning models to determine if the client will subscribe to a term deposit. Accordingly, I use the UCI dataset on bank marketing. Indeed, this dataset is related to the direct marketing campaigns of a Portuguese banking bank. Among trained models, both the Gradient Boosting Classifier (GBC) and Random Forest Classifier (RFC) algorithms deliver the highest score of prediction. The Jupiter code of this project for reference can be found at my repository on the Github space below.