<p>Advanced Combustion Techniques (ACT’s) such as Homogeneous Charge Compression Ignition (HCCI), Partially Premixed Compression Ignition (PCCI), and Reactivity Controlled Compression Ignition (RCCI) have emerged as promising solutions to achieve cleaner and efficient combustion in internal combustion engines (ICE’s). These strategies offer reductions in nitrogen oxides (NOₓ) and soot emissions while enhancing thermal efficiency. However, their practical implementation is constrained by limited combustion phasing controllability, narrow stable operating ranges, strong sensitivity to operating conditions, and difficulties associated with transient engine control. These challenges arise from highly nonlinear interactions between fuel reactivity, injection strategies, air–fuel mixing, and in-cylinder thermochemical processes, which significantly limit the effectiveness of conventional physics-based or rule-based control approaches. Machine learning (ML) has emerged as a promising tool to model and support complex combustion behaviour, prediction, optimization, diagnostics, and control-oriented decision making. This review examines the application of ML techniques in diesel-based HCCI, PCCI, and RCCI engines, with a specific focus on control-relevant challenges rather than performance or emission prediction. The scope of the review is to include recent experimental and ML-based studies across HCCI, PCCI, and RCCI combustion modes which identifies dominant algorithms, key limitations related to robustness and data dependency, and gaps in real-time applicability. These insights aim to guide the development of practical ML-assisted advanced combustion control strategies for reliable operation under dynamic engine conditions.</p>