本文選題：注意 + 強(qiáng)化學(xué)習(xí)　； 參考：《應(yīng)用心理學(xué)》2017年01期

【摘要】：本文在分析總結(jié)現(xiàn)有注意理論的基礎(chǔ)上,假設(shè)注意是一種信息選擇現(xiàn)象,而非心理結(jié)構(gòu)或資源。通過(guò)借鑒人工智能領(lǐng)域強(qiáng)化學(xué)習(xí)算法的思想,筆者提出了一種可以表現(xiàn)出注意現(xiàn)象的人類(lèi)強(qiáng)化學(xué)習(xí)模型。該模型描述了人與環(huán)境交互的過(guò)程:人接受環(huán)境的反饋,根據(jù)自身心理狀態(tài)調(diào)整行為策略,以最大化所獲收益。該過(guò)程中,注意體現(xiàn)為高價(jià)值信息逐漸獲得優(yōu)先加工的現(xiàn)象。因此,本文對(duì)注意的本質(zhì)進(jìn)行了重新思考,為未來(lái)注意研究提供了新思路。
[Abstract]:On the basis of analyzing and summarizing the existing attention theory, this paper assumes that attention is a phenomenon of information selection, not a psychological structure or resource.By learning from the idea of reinforcement learning algorithm in artificial intelligence field, the author proposes a human reinforcement learning model which can show attention phenomenon.The model describes the process of interaction between human and environment: people receive feedback from the environment and adjust their behavior strategies according to their psychological state to maximize the benefits.In this process, attention is reflected in the phenomenon that high-value information gradually obtains priority processing.Therefore, this paper reconsiders the essence of attention and provides new ideas for future attention research.
【作者單位】： 浙江大學(xué)心理與行為科學(xué)系;
【基金】：國(guó)家自然科學(xué)基金項(xiàng)目(31571119,31600881,61431015) 中央高�；究蒲袠I(yè)務(wù)費(fèi)專(zhuān)項(xiàng)資金資助
【分類(lèi)號(hào)】：B842.3
，

本文編號(hào)：1739198