【摘要】：本文通過(guò)對(duì)強(qiáng)個(gè)體效應(yīng)近似因子模型ER方法的再理解,嘗試?yán)脵C(jī)器學(xué)習(xí)方法對(duì)ER法進(jìn)行改進(jìn),嘗試尋找其改進(jìn)算法解決ER方法在強(qiáng)個(gè)體情況下失效的情況,并與已經(jīng)提出的利用有界單調(diào)映射方法進(jìn)行比較,得到了在改進(jìn)估計(jì)結(jié)果的同時(shí)對(duì)個(gè)體效應(yīng)強(qiáng)度識(shí)別能力較強(qiáng)的因子個(gè)數(shù)估計(jì)方法。在文章的敘述過(guò)程中,不僅提供解決方法,還以解決問(wèn)題的思路為順序進(jìn)行敘述,并關(guān)注算法的實(shí)現(xiàn)與優(yōu)化。
[Abstract]:In this paper, we try to improve the ER method by using machine learning method through the reunderstanding of the ER method of the strong individual effect approximate factor model, and try to find its improved algorithm to solve the failure of the ER method in the strong individual case. Compared with the proposed method of bounded monotone mapping, a new method for estimating the number of factors is obtained, which not only improves the estimation results, but also has a better ability to identify the individual effect intensity. In the course of the narration, not only the solution method is provided, but also the idea of solving the problem is described in order, and the realization and optimization of the algorithm are paid attention to.
【學(xué)位授予單位】：浙江工商大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP181
，

本文編號(hào)：2217689