Short answer: Yes, you can.

Predicting price movements in financial markets has always been a challenging task due to the complex, noisy, and often non-linear nature of market data. However, advances in machine learning (ML) and artificial intelligence (AI) have opened new avenues for forecasting prices with increasing accuracy. Numerous academic studies and practical implementations demonstrate that ML techniques, especially when combined with innovative data representations, can provide valuable predictive power.

The State of the Art: Hybrid Methods Combining Complex Networks and AI
One particularly promising approach is detailed in the paper:
https://www.sciencedirect.com/science/article/abs/pii/S0306261918304860
„A novel hybrid method of forecasting crude oil prices using complex network science and artificial intelligence algorithms”
Authors: Minggang Wang, Longfeng Zhao, Ruijin Du, Chao Wang, Lin Chen, Lixin Tian, H. Eugene Stanley
(Available at: ScienceDirect)

What is this method about?
The authors propose a hybrid forecasting model that leverages:

Complex Network Science: Financial price series are transformed into networks that capture the structural patterns and relationships within the data. This network representation can reveal hidden dynamics and dependencies that traditional time series analysis might miss.

Artificial Intelligence Algorithms: Once the price data is mapped into a network structure, AI models such as neural networks or ensemble methods are applied to extract predictive features and forecast future price movements.

This combination allows the model to capture both the temporal dynamics and the intricate structural properties of price fluctuations, leading to improved forecasting performance.

Applicability Beyond Crude Oil
While the study focuses on crude oil prices, the methodology is versatile and can be applied to a broad range of financial instruments, including:

Stocks: Equity prices often exhibit complex interdependencies and patterns that can be captured by network-based models.

Cryptocurrencies: Given their high volatility and unique market dynamics, crypto assets are prime candidates for advanced ML forecasting.

Commodities: Similar to crude oil, other commodities like metals, agricultural products, and energy resources can benefit from these techniques.

The key lies in transforming raw price data into a meaningful representation (e.g., a network) that AI algorithms can effectively analyze.

Alternative Approaches
Another interesting perspective on this topic is presented in the paper:
https://www.mdpi.com/2227-7390/10/22/4361
„Forecasting Financial Time Series with Machine Learning: A Review”
(Available at: MDPI)

This review highlights various ML models used in financial forecasting, including:

Support Vector Machines (SVM)

Random Forests

Long Short-Term Memory networks (LSTM)

Convolutional Neural Networks (CNN)

It emphasizes the importance of feature engineering, data preprocessing, and model selection tailored to the specific characteristics of financial time series.

What to Do with Predicted Prices?
The practical use of predicted prices depends on your goals:

For Traders:
Use forecasts to inform entry and exit points, risk management, and portfolio allocation. Machine learning can help identify short-term price trends or reversals, potentially increasing trading profitability.

For Investors:
Use predictions to time purchases or sales of assets, optimize asset allocation, or hedge risks.

For Currency Buyers:
If you are buying dollars or other currencies, forecasts can guide your timing decisions, helping you avoid unfavorable rates.

Regardless of your role, it is crucial to combine ML predictions with sound risk management and domain knowledge.

How I Apply These Methods
To see a practical demonstration of applying these concepts, watch my walkthrough here:

YouTube Video: How I Use Machine Learning for Price Prediction

In this video, I show step-by-step how to:

Transform price data into network representations

Train AI models on these features

Interpret and use the predictions in trading decisions

Final Thoughts
Machine learning is not a crystal ball – no model can predict markets with 100% accuracy. However, by leveraging advanced techniques like complex network analysis combined with AI, you can gain a statistical edge and better understand market dynamics.

If you are a trader or investor interested in modern quantitative methods, I highly recommend diving deeper into this exciting field. Start experimenting with your own data, explore various ML models, and continuously refine your approach.