Fusarium wilt of tomato, caused by Fusarium oxysporum f. sp. lycopersici (FOL), severely impacts global tomato production. A field experiment evaluated 25 treatments, categorised as compost-based [Farmyard manure (FYM) + de-oiled sweet basil compost], biological control (FYM + bio-agents) and chemical fungicides under field conditions. A random forest machine learning (ML) model predicted treatment effects, validated through biological assays and statistical tests. The treatment combining FYM + de-oiled sweet basil compost (25 t/ha) with Trichoderma viride (5 kg/ha) and Pseudomonas fluorescens (5 kg/ha) demonstrated the highest efficacy. It reduced disease incidence while maintaining soil health. The model revealed antagonism between bio-agents and fungicides. This study highlights ML's potential in optimising integrated disease management (IDM), reducing resource use and expediting decision-making. The validated model provides a reliable decision-support tool for precise and sustainable management of Fusarium wilt in tomato.