Advanced Gravitational-Wave detectors like LIGO aim to detect gravitational waves by laser interferometry. LIGO is currently the most sensitive instrument with the capability to detect a change in distance of the order of 10-20 m. To increase the sensitivity and frequency of detection of gravitational waves, there is a need to suppress the various forms of noise that hinder the operation of the detector. Our work focuses on developing techniques to reduce Newtonian noise, seismic noise and laser phase noise that affect terrestrial Gravitational-Wave detectors. We address the problem of finding an optimal sensor array configuration for broadband Newtonian noise cancellation at LIGO by formulating a suitable multi-objective optimization problem. We propose two solution approaches. The first approach obtains a smart Pareto optimal solution by filtering a set of Pareto points generated by the Normal Boundary Intersection method. The second approach utilizes the Reinforcement Learning framework to solve the multi-objective optimization problem and directly obtain a smart Pareto solution. The proposed methods generate solutions close to the sensor noise limit. To reduce seismic noise at LIGO, we propose a Reinforcement Learning-based local control of the multi-stage pendulum suspension system. This seminar also presents a preliminary control setup to modulate phase fluctuations in the LIGO phase noise measurement system