Week #2 #

Detailed Requirements Elaboration #

This week, the team focused on clarifying the essential analytical and technical requirements for the MVP, with particular emphasis on the machine learning component. The main efforts were directed toward time series analysis and meaningful feature selection based on available gold market data.

analyzed short-term gold distribution similarity: a statistical comparison of gold price distributions in Russian and U.S. markets was conducted. It was found that both follow a relatively similar trend structure.
time series analysis: volume distribution analysis, open and close price similarity, high and close as well as low and close price similarities similarities, stationarity hypothesis testing for close price, autocorrelation and partial autocorrelation visualization.
feature selection for short-term forecast: these features include historical prices, market-derived metrics etc.

Based on early findings, the team decided to priotize short-term forecasting of gold prices using yechnical indicators. This will serve as primary functionallity of the MVP.

Prioritized backlog #

Kanban board

Project specific progress #

Frontend #

UI design created for FHD monitors.
HTML structure is set.
CSS is configured.
Header section implemented.

Backend #

Possible API endpoints written.
API structure set.

ML #

Relevant datasets identified and collected.
Indicators selected and justified (incl. volatility, market filters).
Time series properties analyzed.
Baseline model for gold forecasting implemented.
External data sources (e.g. Kaggle) reviewed.
Early preprocessing and EDA completed.
Hypothesis testing for distribution similarities between time-series.
Review of modern forecasting approaches to select model.
MVP and project-vision models selection.

Weekly commitments #

Individual contribution of each participant #

Vladimir Toporkov - Wrote second report, created the initial HTML structure for the frontend and outlined the basic CSS classes for future styling. Implemented header section. Commit link
Farit Sharafutdinov - Made justifications of the choice of indicators and distribution of the Russian and American markets similarities (showed minor statistical differences in a short period, according to Kolmogorov-Smirnov).
Ilya Grigorev - Baseline model implemented, selected a model for MVP, continued the analysis of the characteristics of the gold price time series ( Link to the notebook).
Rail Sharipov - Made feature analysis for short-term and long-term forecast. Link
Askar Kadyrgulov - Created Kanban board. Implemented some endpoints.
Nikita Solomennikov - Created the first version of the design in Figma. Link to the commit and Link to figma

Plan for Next Week #

Create UI-kit: start working with animations and color namings.
Finish FHD frontend part: complete main section and implement UI-kit.
Implement API: integrate the backend endpoints with the frontend
Select features for long-term forecast: analyze available datasets and select relevant features that impact long-term trends.
Deploy the first version of the model: package the current machine learning model and deploy it
Finish with feature selection fot short-term forecast: finalize the preprocessing pipeline for short-term prediction.

Confirmation of the code’s operability #

We confirm that the code in the main branch:

[✓] In working condition.
[✓] Run via docker-compose (or another alternative described in the README.md).