This semester I am taking ECE EL-9343 : Data Structure and Algorithm. And I would draw parallel between the creative thinking method we use and those computer scientists use.
Greedy Algorithm and Mind Mapping:
The composition of mind mapping is parallel to greedy algorithm. The point of mind mapping is all about estimating the principle dimension which an object/concept lies on linguistically. Doing a lexicon analysis is way too expensive so we are using mind mapping to grab a general out look.
The central idea of greedy is that we do not consider the problem comprehensively but locally. As long as the problem could be consider non-caustic, it would be fine for us to pile up the local optimization and assume it is a global optimization. The problem of mind mapping is that it is such a crude estimation. The builder would generally desire a quick result. In this way, we might overlook the connection between different nodes and regard them independently. So the building of mind map is generally a greedy process.
Free Writing and q-Learning:
Free Writing itself is an typical trial-and-error learning method. It is easily connected to reinforcement learning method such as q-learning. The central idea of q-learning is that we try randomly at first to figure out a q-table, which indicates a function determines the relationship between current state, expected actions and the reward value.
Similar to free writing, the initial value of q-learning is random, and we would not judge the previous action. Instead, based on previous action and result, we update the expectation table for a potential wiser choice next time.
Comments