A CSE449 distributed-systems project. A master node loads historical temperature records from GodaGari–Amtoli in Rajshahi and hands training data to worker machines over sockets: more to the ones with room to spare, less to the busy ones, going by the CPU cores, memory and live load each worker reports. Workers train regression models (linear regression, random forest, SVR) on their share and send them back, and the master combines them, evaluates the result and plots it.
