Algorithmic Trading and Short-term Forecast for Financial Time Series with Machine Learning Models Implementation and Testing Workshop

---

The increase in popularity of machine learning models and algorithms has brought up methods for increasing the efficiency of processes in various industries and new use cases. From corporations to individuals, machine learning is applicable in a wide range of settings. One such application is using machine learning models for stock price forecasting. This area is often studied. However, it still needs to solve problems due to the high complexity and volatility that technical factors and sentiment-analysis models try to capture. Based on this premise, we suggested building up our previous experience building short-term forecasting models using machine learning models. The research project aims to develop practical algorithmic trading algorithms based on accurate short-term forecasts for financial time series using machine learning models. The project focuses on these research activities:

1. Building a database containing the targeted financial and other time series is valuable input to the machine learning models. This streamlines the process of acquiring financial data required for training the machine learning model by adding a web UI for configuring data collection automation and extracting required data.
2. A short-term forecasting model based on machine learning is developed using the various time-series data from the data warehouse. Machine learning algorithms such as neural networks, random forecasts, support vector regression, XGBoost, and long short-term memory (LSTM) are evaluated with several performance criteria to identify the most accurate model from a short-term trading perspective. After the performance analysis based on several metrics, XGBoost was chosen for further development.

We will demonstrate how to evaluate the quality of the collected data, the metrics of the machine learning model’s prediction, and their integration. 

---

---

---

---

1. Welcome from the IEEE Okanagan Subsection, IEEE Okanagan College Student Branch and Computer Science Department: Dr. Youry Khmelevsky, Chair (9:30 am – 9:35 am)
2. Invited speaker from the University of Paris East Creteil: Dr. Gaétan Hains (9:35 am – 10:15 am)
3. Invited speakers from Langara College: Dr. Albert Wong & Andres C. Viloria Garcia (10:15 am – 11:00 am)
      Machine Learning algorithms for stock price forecasting
      Model building process and methodology
      Normalization and performance evaluation
      Iterations and results
4. Coffee Break (11:00 am - 11:15 am)
5. Student research project presentation and demonstration (11:15 am - 12:30 pm)
      Database design
      Data collection process automation
      Extraction, Transformation and Loading Process
      UI development for data configurations and extraction
      Integration prototype using Jupyter notebooks

We will demonstrate how to evaluate the quality of the collected data, the metrics of the machine learning model’s prediction, and their integration. 

---