In the article, ML-based trends following techniques are used to include machine learning (ML) models in trading and investing. Old-school trend-following algorithms like Ichimoku Cloud, get complemented by cutting-edge ML methods such as Naive Bayes, K-Nearest Neighbors, Gradient Boosting, Decision Trees, and so on. They are backtested using currency pairs (2010–2023) data, and their relative performance against parameters such as total return, maximum drawdown, and mean return. Data indicates that Extreme Gradient Boost (XGB) or Naive Bayes algorithms give far better returns at low risk than the standard approach. Additionally, time series models. as TCNs, and Kalman Filters) for accurate predictions. Both models have major upsides on total returns and mean return with lower volatility. The paper describes how machine learning and time series models can be used to improve decision making in the financial markets and increase the profitability of trend-following strategies.