The Skill-Task Matching Model: Mechanism, Model Form and implications

We propose the Skill-Task Matching Model as a supplement to the Cobb-Douglas production function. We hold that firms set expected profits in the beginning of each producing period, then try to achieve them. The ability to achieve target number is not born. Firms continuously trial and error, make their actual profits increasingly approach expected profits. To this iteration process, the model vectorizes occupations into a task vector space, vectorizes employees into a skill vector space, regards production technology as the matching level of skills and tasks, and regards corporate strategy as the value vector of tasks. Firms turn these producing parameters by iteration, to make actual profit get closer to expected profit. We build a neural network algorithm to simulated how the Skill-Task Matching Model works. Being a new model to illustrate firm's decision making, the Skill-Task Matching Model offers another perspective to discuss firm's short-run and long-run decision making, routine tasks, and perfect competition markets.
Comment: 23 pages, 7 figures