The orbitofrontal cortex (OFC) is found to be one of the crucial regions in the prefrontal cortex (PFC) involved in decision-making and instrumental learning, influencing the downstream processes in the subcortical regions. However, the exact mechanisms through which the OFC encodes, or processes information related to decision-making and learning are unclear. Specifically, several lesion studies have shown that two sub-regions of the OFC – lateral OFC and the medial OFC (ventromedial PFC) – have differential roles in the process of instrumental learning in the context of multi-alternative choice. Here, we propose a multi-alternative decision-making model with reward-based learning using evidence accumulation models (EAMs) of decision making and reinforcement learning (RL) models of error-driven learning. More specifically, we use a variation of drift diffusion models (DDM) based on an Advantage Race Diffusion (ARD) framework for choices among two or more options, that captures stimulus difficulty, speed-accuracy trade-off, and stimulus-response-mapping reversal effects (RL-ARD). It is specifically on these aspects of instrumental learning that lateral and medial OFC have been shown to play complementary roles. Unlike DDMs which strongly depend on the accumulation of the evidence difference alone, the ARD based accumulators are driven by both difference and sum of the mean reward expectancies. We studied if the differential contribution of sum and difference of reward expectancies in the RL-ARD model can help explain the roles of lateral and medial OFC in the process of decision-making and learning. We tested the model on a 3-arm bandit task with varying reward probabilities. We simulated lesions of lateral and medial OFC structures in the model and showed differential effects in performance depending on the value difference of the presented options, thus accounting for recent experimental findings about the dissociate functional roles of both the sub-regions.