【正文】 要求積極尋求新的指標(biāo)去衡量新的競(jìng)爭(zhēng)環(huán)境下成功驅(qū)動(dòng)因素，即企業(yè)評(píng)價(jià)體系的重新構(gòu)建。企業(yè)評(píng)價(jià)對(duì)企業(yè)的成敗更具決定意義。西安**大學(xué)本科畢業(yè)設(shè)計(jì)(論文) 1第 一 章 緒 論 企 業(yè) 評(píng) 價(jià) 的 目 的 與 意 義企業(yè)評(píng)價(jià)以一種測(cè)評(píng)制度存在于企業(yè)管理系統(tǒng)之中，是企業(yè)進(jìn)行自我診斷與自我提升的有效手段。它的優(yōu)勢(shì)在于用線性學(xué)習(xí)算法來(lái)完成以往的非線性算法所做的工作，同時(shí)又能保持非線性算法的高度精確性。1988 年，Moody 和 Darken 提出了一種神經(jīng)網(wǎng)路結(jié)構(gòu)，即 RBF 神經(jīng)網(wǎng)絡(luò)，屬于前向神經(jīng)網(wǎng)絡(luò)類型，它能夠以任意精度逼近任意連續(xù)函數(shù)，特別適合解決分類問(wèn)題 [3]。神經(jīng)網(wǎng)絡(luò)計(jì)算技術(shù)目前存在的主要問(wèn)題，簡(jiǎn)而言之，即目前神經(jīng)網(wǎng)絡(luò)計(jì)算技術(shù)還不能有效地解決大規(guī)模的實(shí)際問(wèn)題 [2]。該技術(shù)在許多領(lǐng)域得到應(yīng)用，如智能控制、模式識(shí)別等。神經(jīng)網(wǎng)絡(luò)計(jì)算技術(shù)研究開(kāi)始復(fù)蘇，于是掀起了第二次研究高潮，人們又開(kāi)始躍躍欲試、信心百倍。因此到了六七十年代，大多數(shù)人對(duì)神經(jīng)網(wǎng)絡(luò)計(jì)算技術(shù)的研究熱情大大下降，形成了神經(jīng)網(wǎng)絡(luò)計(jì)算技術(shù)研究的第一次低潮。 神經(jīng)網(wǎng)絡(luò)理論在初期取得了一些進(jìn)展，比如說(shuō)建立了一些人工神經(jīng)網(wǎng)絡(luò)的計(jì)算模型，如前饋網(wǎng)、反饋網(wǎng)等等。人工神經(jīng)網(wǎng)絡(luò)模擬人類部分形象思維的能力是模擬人工智能的一條途徑，特別是可以利用人工神經(jīng)網(wǎng)絡(luò)解決人工智能研究中所遇到的一些難題。關(guān)鍵詞：評(píng)價(jià)方法， PNN 神經(jīng)網(wǎng)絡(luò)，SPSS ，MATLAB西安**大學(xué)本科畢業(yè)設(shè)計(jì)(論文) IABSTRACTEnterprise’s evaluation is an important means to inspect the business’s results and check the developmental direction . How to use quantitative method to evaluate enterprises accurately is a question which people have been researched all along, and the correct evaluation of enterprises can guide our investment decisions. Probabilistic neural work is a which can be used for the this paper, the probabilistic neural work is used to evaluate enterprises,and enterprises are divided into the good and the bad . This paper summarizes the methods which are applied to enterprise’s evaluation, and introduces the probabilistic neural work principle and structure, then probabilistic neural work is applied to evaluation of enterprises. First of all, four among twelve financial indicators of the 60 listed panies are significantly correlated with the yields on corporate bonds by using stepwise regression method. Secondly, four financial indicators are regarded as the input vector and the cases of the good or bad enterprises as the output vector. PNN neural work is established and trained by 60 listed data. Finally, the train samples and test samples are simulated, the results show that the probabilistic neural work is a very effective method to classificate and evaluate enterprise. In closing remarks, the content is summarized and further work is prospected in the future.KEY WORDS:: evaluation method, probabilistic neural work, SPSS, MATLAB西安**大學(xué)本科畢業(yè)設(shè)計(jì)(論文) II目 錄前 言 .....................................................................................................................................................1第 一 章 緒論 .....................................................................................................................................2 企業(yè)評(píng)價(jià)的目的與意義 ...........................................................................................................2 企業(yè)評(píng)價(jià)方法 ...........................................................................................................................2 定性評(píng)價(jià)法 .........................................................................................................................2 主成分分析法 .....................................................................................................................2 模糊綜合 評(píng)價(jià)法 .................................................................................................................3 綜合指數(shù) 法 .........................................................................................................................4 灰色關(guān)聯(lián) 聚類 法 .................................................................................................................5 TOPSIS 法 ...........................................................................................................................6 數(shù)據(jù)包絡(luò) 分析法 .................................................................................................................6 功效系數(shù)法 .........................................................................................................................7 線性加權(quán)綜合法 .................................................................................................................8 層次分析法 .....................................................................................................................10 BP 神經(jīng)網(wǎng)絡(luò)評(píng)價(jià)法 .......................................................................................................10 本 章 小 結(jié) .............................................................................................................................11第 2 章 概率神經(jīng)網(wǎng)絡(luò)（PNN）概述 ...........................................................................................12 概率神經(jīng)網(wǎng)絡(luò)簡(jiǎn)介 .................................................................................................................12 概率神經(jīng)網(wǎng)絡(luò)（PNN）原理 .................................................................................................13 模式分類的貝葉斯判定策略 ...........................................................................................13 密度估計(jì)的一致性 ...........................................................................................................14 概率神經(jīng)網(wǎng)絡(luò)（PNN）模型 ...........................................................................................15第 3 章 概率神經(jīng)網(wǎng)絡(luò)（PNN）在企業(yè)評(píng)價(jià)中的應(yīng)用 ...............................................................17西安**大學(xué)本科畢業(yè)設(shè)計(jì)(論文) III 樣本的選取與確定 .................................................................................................................17 指標(biāo)的選取與確定 .................................................................................................................17 PNN 網(wǎng)絡(luò)模 型仿真 ................................................................................................................19 網(wǎng)絡(luò)層數(shù)的確定 ...............................................................................................................19 輸入個(gè)數(shù)、輸出層結(jié)點(diǎn)的確定 .......................................................................................19 ...............................................................................................................19 網(wǎng)絡(luò)模型 ...........................................................................................................................19 訓(xùn)練樣本的網(wǎng)絡(luò)仿真 ....