Abstract

Megatrends are long-term but also massive moves. Proper identification of the
megatrend in investing can lead to good above-average profits.

Time-series data are difficult for machine learning. There is a question of
what is the best representation of them. Our approach is using the correlation
coefficient as a metric for the similarity of two series.

Finally, clustering is used for grouping tokens. Analysis of cluster trends is
a potential tool for investors' decisions.

Prezentace

Záznam

Readings

[Time-series clustering – A decade review]
(https://www.sciencedirect.com/science/article/pii/S0306437915000733)

[Financial Times Series Forecasting of Clustered Stocks]
(https://link.springer.com/article/10.1007/s11036-020-01647-8)

Optional:

[Clustering and Regression Techniques for Stock Prediction](https://www.sciencedirect.com/science/article/pii/S2212017316301931)