201232433 六、發明說明： 【發明所屬之技術領域】 本發明是有關於-種自動化績效評估方法及系統，特 別是指-種運用學習曲線理論與迴歸分析之自動化績效評 估方法及系統。 【先前技術】 參閱圖卜如圖1學習成長曲線所示，其顯示會隨時間 的增加而提昇每單位的工作績效，且此 型的成長曲線。當扣除一開始的啟蒙暖身期(如丨線至為二 ，可想像有條虛擬成長迴歸直線穿過該s曲線。此虛擬的 迴歸平均值可以做為企業内部評估個人績效的一項『效果 量標準』。 -般來說’管理階層會期望員工能夠愈早成長愈好(即 緩慢開始的期間需愈短愈好），也期望員工能夠愈加速成長 愈子（ί7心遽加速的曲線需愈陡愈好），因此管理階層期望員 工的總工作效果等於愈早成長效果及愈速成長效果之合， 而如何找出一種能夠證釋愈早愈陡效果之演算法，就成為 將學I曲線與迴歸分析結合以評估員工績效所面臨的課題 〇 【發明内容】 因此，本發明之目的，即在提供一種自動化績效評估 方法。 於疋’本發明自動化績效評估方法包含：（Α)提供一實 際工作量資料庫，該實際工作量資料庫包括一受評估人員 201232433 在多個時間區間的一組工作量數值；（B)運算與該組工作量 數值相對應之一組工作量迴歸值、一相關係數，以及一 R 平方值’（c)根據該組工作量數值、該組工作量迴歸值，以 及該相關係數，運算一期望總效果值；（D)根據該期望總效 果值、相關係數，以及R平方值，運算一調整後總效果值 ，以及（E)根據該調整後總效果值以及一預定效果係數，運 算該受評估人員之一績效值。 本發明之另一目的’即在提供一種自動化績效評估系 統。 於是，本發明自動化績效評估系統適用於自動評估一 受評估人員之工作績效。該自動化績效評估系統包含一實 際工作量資料庫、一迴歸數據運算模組、一期望總效果值 運算模組、一調整後總效果值運算模組及一績效值運算模 組。該貫際工作量資料庫包括該受評估人員在多個時間區 間的一組工作量數值。該迴歸數據運算模組用以運算與該 組工作量數值相對應之一組工作量迴歸值、一相關係數， 以及一 R平方值。該期望總效果值運算模組用以根據該組 工作量數值、該組工作量迴歸值，以及該相關係數，運算 期望總效果值。該調整後總效果值運算模組用以根據該 期望總效果值、相關係數，以及R平方值，運算一調整後 總效果值。該績效值運算模組用以根據該調整後總效果值 以及一預定效果係數，運算該受評估人員之一績效值。 本發明之功效在於，可將受評估人員的工作量數值自 動轉換為績效值，藉以對早成長以及陡成長者予以鼓勵。 201232433 【實施方式】 有關本發明之前述及其他技術内容、特點與功效，在 以下配合參考圖式之一個較佳實施例的詳細說明中將可 清楚的呈現。 在本發明被詳細描述之前，要注意的是，在以下的說 明内容中，類似的元件是以相同的編號來表示。201232433 VI. Description of the invention: [Technical field to which the invention pertains] The present invention relates to an automated performance evaluation method and system, and in particular to an automated performance evaluation method and system using a learning curve theory and regression analysis. [Prior Art] Referring to Figure 2, the growth curve is shown in Figure 1. The display shows that the performance per unit will increase with time, and this type of growth curve. When deducting the beginning of the enlightenment warm-up period (such as the 丨 line to two, you can imagine a virtual growth regression line through the s curve. This virtual regression average can be used as an effect in the internal evaluation of individual performance The standard of quantity. - Generally speaking, the management team expects employees to grow as early as possible (that is, the shorter the period is needed, the shorter the better), and the more the employees are expected to accelerate their growth. The steeper the better, so the management expects that the total work effect of the employees is equal to the combination of the earlier growth effect and the faster growth effect. How to find an algorithm that can prove the effect of the earlier and steeper, becomes the learning I Curves and regression analysis combined to assess the problems faced by employee performance 发明 [Summary of the Invention] Therefore, the object of the present invention is to provide an automated performance evaluation method. The invention's automated performance evaluation method includes: (Α) provides a The actual workload database, the actual workload database includes a set of workload values of the evaluator 201232433 in multiple time intervals; (B) operation The set of workload values corresponds to a set of workload regression values, a correlation coefficient, and an R-squared value '(c) based on the set of workload values, the set of workload regression values, and the correlation coefficient, an expectation a total effect value; (D) calculating an adjusted total effect value based on the expected total effect value, the correlation coefficient, and the R square value, and (E) calculating the received value based on the adjusted total effect value and a predetermined effect coefficient One of the evaluation personnel's performance values. Another object of the present invention is to provide an automated performance evaluation system. Thus, the automated performance evaluation system of the present invention is adapted to automatically evaluate the performance of an evaluator. The automated performance evaluation system includes a The actual workload database, a regression data operation module, a desired total effect value calculation module, an adjusted total effect value calculation module, and a performance value calculation module. The continuous workload database includes the evaluated A set of workload values for a plurality of time intervals. The regression data operation module is used to calculate a value corresponding to the workload value of the group. a set of workload regression values, a correlation coefficient, and an R-squared value. The expected total effect value calculation module is configured to calculate a desired total effect value according to the set of workload values, the set of workload regression values, and the correlation coefficient The adjusted total effect value calculation module is configured to calculate an adjusted total effect value according to the expected total effect value, the correlation coefficient, and the R square value. The performance value calculation module is configured to use the adjusted total effect value according to the adjusted total effect value. And calculating a performance value of the one of the evaluated persons according to a predetermined effect coefficient. The effect of the invention is that the workload value of the evaluator can be automatically converted into a performance value, thereby encouraging early growth and steep growth. 201232433 [ The above and other technical contents, features, and advantages of the present invention will be apparent from the following detailed description of the preferred embodiments of the invention. In the following description, like elements are denoted by the same reference numerals.
參閲圖2’本發明自動化績效評估系統丨之較佳實施例 適用於自動評估一受評估人員之工作績效。該自動化績效 評估系統1包含-實際卫作量資料庫U、—迴歸數據運算 模=12、m絲值運算模組13、—調整後總效果值 運算模組】4，以及一績效值運算模組15。 該實際工作量資料庫n包括該受評估人員在多個時間 區間的-組卫作量數值。在以下實施例中，係假設該受評 估人員為醫院中的f師’該等時間區間是指連續的數週， 且該組工作量數值是指此連續數週的每週本診次數，例如 以下將以某醫師連續33週的每週本診次數來做為範例，來 說明本發明如何根據受評估人員的工作量，自動運算出其 績效值（如應獎勵或懲罰的金額）。 /、 該迴歸數據運算模組12 _運算與触4量數值相 子心、之、.且工作里迴歸值、一相關係數，以及一統計學上 用以解釋變異量的R平方。 1該期望總效果值運算模組13用以根據該組卫作量數值 垓組工作罝迴歸值，以及該相關係數，運算一期望 果值。 201232433 該調整後總效果值運算模組14用以根據該期望總效果 值、相關係數’以及R平方值，運算一調整後總效果值。 該績效值運算模組15用以根據該調整後總效果值以及 一預疋效果係數，運算該受評估人員之一績效值。 參閱圖3 ’以下將以某醫院某醫師連續33週之每週門 診次數做為實際卫作量資料庫11中的工作量數值資料，來 :月本發明自動化績效評估方法之較佳實施例如何從該組 作量數值，自動運算出該醫師之績效值。首先，如圖4 步驟21所示，提供管 ,a '、貫際工作量資料庫11’其包括需受評估 人員在該等連續33 加 _ 週的—組工作量數值，如以下表1第2 斕所示。 表1Referring to Figure 2', a preferred embodiment of the automated performance assessment system of the present invention is adapted to automatically assess the performance of an evaluator. The automated performance evaluation system 1 includes an actual workload database U, a regression data operation module = 12, a m-wire calculation module 13, an adjusted total effect value calculation module 4, and a performance value calculation module. Group 15. The actual workload database n includes the number of the group's workload in the plurality of time intervals. In the following embodiments, it is assumed that the evaluator is a teacher in the hospital. The time intervals refer to consecutive weeks, and the workload value refers to the number of weekly clinics for the consecutive weeks, for example. The following is an example of the number of weekly visits by a physician for 33 consecutive weeks to illustrate how the present invention automatically calculates its performance value (such as the amount of reward or penalty) based on the workload of the assessee. /, the regression data operation module 12 _ operation and touch 4 amount of values, the center of the heart, and the working regression value, a correlation coefficient, and a statistically used to explain the R square of the variation. 1 The expected total effect value calculation module 13 is configured to calculate a desired fruit value according to the set of workload values, the group work 罝 regression value, and the correlation coefficient. 201232433 The adjusted total effect value calculation module 14 is configured to calculate an adjusted total effect value according to the expected total effect value, the correlation coefficient ', and the R square value. The performance value calculation module 15 is configured to calculate a performance value of the one of the evaluated persons according to the adjusted total effect value and a preview effect coefficient. Referring to Figure 3, the following is a summary of the number of weekly outpatient visits by a physician in a hospital for 33 consecutive weeks as the workload data in the actual workload database 11 to: How is the preferred embodiment of the monthly automated performance evaluation method of the present invention? From the set of values, the physician's performance value is automatically calculated. First, as shown in step 21 of FIG. 4, a tube, a ', and a continuous workload database 11' is provided, which includes the workload value of the group to be subjected to the continuous 33 plus _ weeks, as shown in Table 1 below. 2 is shown. Table 1
S 6 201232433S 6 201232433
10 80 70 70 11 63 72 72 12 78 75 75 13 83 77 77 14 78 79 79 15 86 81 81 16 99 83 83 17 77 85 85 18 87 87 87 19 92 89 89 20 95 91 91 21 21 93 93 22 77 95 95 23 79 97 97 24 83 99 99 25 75 101 101 26 72 103 103 27 120 105 105 28 108 107 107 201232433 29 124 109 109 30 157 111 111 31 191 113 113 32 80 115 115 33 115 117 117 總和=2796 總和=2760 接著，如步驟22所示，該迴歸數據運算模組12運算 與該組工作量數值相對應之一組工作量迴歸值（如表1第3 欄所示）及一 R平方值，並自動繪製成迴歸曲線統計圖表， 如圖4所示。圖4中除了繪示該組工作量數值曲線以及工 作量迴歸值迴歸直線外，也將其迴歸直線方程式 y=2.0424x+50.006以及R平方值r2=〇 3442顯示在右上角。 另外，該迴歸數據運算模組12也同時運算了相關係數=〇 59 ’以供用於本發明方法後續的運算過程。10 80 70 70 11 63 72 72 12 78 75 75 13 83 77 77 14 78 79 79 15 86 81 81 16 99 83 83 17 77 85 85 18 87 87 87 19 92 89 89 20 95 91 91 21 21 93 93 22 77 95 95 23 79 97 97 24 83 99 99 25 75 101 101 26 72 103 103 27 120 105 105 28 108 107 107 201232433 29 124 109 109 30 157 111 111 31 191 113 113 32 80 115 115 33 115 117 117 Sum = 2796 Sum=2760 Next, as shown in step 22, the regression data operation module 12 calculates a set of workload regression values corresponding to the set of workload values (as shown in the third column of Table 1) and an R-squared value, And automatically drawn into a regression curve statistical chart, as shown in Figure 4. In Fig. 4, in addition to the numerical value curve of the workload and the regression line of the regression value of the workload, the regression equation y=2.0424x+50.006 and the R-squared value r2=〇 3442 are also displayed in the upper right corner. In addition, the regression data operation module 12 also computes the correlation coefficient = 〇 59 ' for the subsequent operation of the method of the present invention.
接著，如步驟23所示，該期望總效果值運算模組13 根據該組工作量數值、該組工作量迴歸值，以及該相關係 數，運算一期望總效果值。在期望總效果值運算模組丨3運 算期望總效果值的過程中，係先判定該相關係數是否大於 零。若該相關係數大於零，表示受評估人員的表現為正成 長，則期望總效果值運算模組13運算期望總效果值=該組 工作量數值之總和一該組工作量迴歸值之總和（即殘差總和） 。反之，若該相關係數小於零，表示受評估人員的表現為 負成長，則期望總效果值運算模組丨3運算期望總效果值L 201232433 —(該組工作量數值之總和減去該組工作量迴歸值之總和之 絕對值）。此外，由於在本實施例中，是假設前2週的觀察 資料不列入計算，故期望總效果值運算模組13在根據表^ 貫際門診篁及工作量迴歸值運算期望總效果值時，事實上 是取第2攔以及第4欄的數據來進行運算。因此，在相關 係數0.59大於零的情況下，期望總效果值=表1第2棚數據 總和2796—表1第4欄數據總和2760=36次門診。 此外，由於期望總效果值為正值並不必然是愈早愈陡 的必然結果，因此在本發明實施例中，係先設定『負成長 必不能為正效果』#消極條件。再者，在本發明實施例中 ，也設定了『愈陡成長的效果權重愈高』的積極條件。故 愈陡的正效果愈會加重其權重。因此，本發明實施例運用 統計學中解釋變異量的R平方來作為愈陡成長的加成權重 。因此，可以導出一個調整後總效果值法則如以下公式 ⑴： " 調整後總效果值=期望總效果值X(1+R平方）（】）。 然而，假設兩位受評估人員的負期望總效果相同則 依據公式（1)，成長較遲者會因R平方較大而遭受較大的懲 罰’於是在本發明實施例中認為大器晚成者較扶不起的阿 斗要好一些的情況下，便可將公式（1)修改為以下公式（2): 調整後總效果值，期望總效果值>0,期望總效果值x (i+R平方），期望總效果值x(1 —R平方））（2)。 /旦是’公式⑺也有個盲點’就是當負成長且為負期望 L效果時，R平方愈大（即負成長愈大）的被懲罰量會小於 201232433 負成長愈小者。因此，為修正此盲點，在本發明較佳實施-例中，便可將公式（1 )、（2)整合成為以下公式（3): - 〃負成長必為負效果 ^正成長時，使用公式 、負成長時，使用公式（2) 因此，接著如步驟24所示，該調整後總效果值運算模 組14根據該期望總效果值、相關係數，以及R平方值，運 算一調整後總效果值。在調整後總效果值運算模組丨4運算 調整後總效果值的過程中，係先判定該相關係數是否大於春 零。若該相關係數大於零，表示受評估人員的表現為正成 長，則調整後總效果值運算模組14進一步視期望總效果值 是否大於零來決定如何調整後總效果值。若該期望總效果 值大於零’則調整後總效果值運算模組14運算該調整後總 效果值=該期望總效果值χ( 1+R平方值）；若該期望總效果值 小於零，則調整後總效果值運算模組14運算該調整後總效 果值=該期望總效果值X(1_R平方值）。 反之，若調整後總效果值運算模組14判定該相關係數鲁 小於零’表示受評估人員的表現為負成長，則調整後總效 果值運算模組14運算該調整後總效果值=該期望總效果值父 |1+砰方値|。 因此’在本實施例相關係數0.5 9大於零且期望總效果 值36次皆大於零的情況下’該調整後總效果值模組14運 算出調整後總效果值=46次門診。 然後，該績效值運算模組15根據該調整後總效果值以Next, as shown in step 23, the expected total effect value calculation module 13 calculates a desired total effect value based on the set of workload values, the set of workload regression values, and the phase correlation number. In the process of expecting the total effect value calculation module 丨3 to calculate the expected total effect value, it is first determined whether the correlation coefficient is greater than zero. If the correlation coefficient is greater than zero, indicating that the evaluated person's performance is positive growth, then the total effect value computing module 13 is expected to calculate the expected total effect value = the sum of the set of workload values and the sum of the set of workload regression values (ie, The sum of the residuals). Conversely, if the correlation coefficient is less than zero, indicating that the performance of the evaluator is negative growth, then the total effect value calculation module 丨3 is expected to calculate the expected total effect value L 201232433 — (the sum of the set of workload values minus the set of work The absolute value of the sum of the regression values). In addition, in the present embodiment, it is assumed that the observation data of the first two weeks is not included in the calculation, so that the total effect value calculation module 13 is expected to calculate the expected total effect value according to the table of the clinic and the workload regression value. In fact, the data of the second and fourth columns are taken for calculation. Therefore, in the case where the correlation coefficient 0.59 is greater than zero, the total effect value is expected = the second shed data of Table 1 is 2796 - the sum of the data in column 4 of Table 1 is 2760 = 36 outpatients. In addition, since the total effect value is expected to be a positive value, it is not necessarily the inevitable result of the earlier and steeper. Therefore, in the embodiment of the present invention, the "negative growth must be a positive effect" #negative condition is first set. Further, in the embodiment of the present invention, the positive condition of "the higher the weight of the effect of the steeper growth" is set. Therefore, the steeper the positive effect will increase its weight. Therefore, the embodiment of the present invention uses the R-square of the variation amount explained in the statistics as the addition weight of the steeper growth. Therefore, you can derive an adjusted total effect value rule such as the following formula (1): " Adjusted total effect value = expected total effect value X (1 + R square) (]). However, assuming that the two respondents have the same negative expected total effect, according to formula (1), those who grow later will suffer greater punishment because of the larger R square. Thus, in the embodiment of the present invention, it is considered that the latecomers are better. If the Abu is better, you can change the formula (1) to the following formula (2): Total effect after adjustment, expected total effect value > 0, expected total effect value x (i + R square) , the total effect value x (1 - R square) is expected (2). / Dan is that 'Formula (7) also has a blind spot'. When negative growth and negative expectation L effect, the greater the R square (ie, the greater the negative growth), the less the penalty will be less than 201232433. Therefore, in order to correct this blind spot, in the preferred embodiment of the present invention, the formulas (1) and (2) can be integrated into the following formula (3): - The negative growth must be a negative effect. In the formula and negative growth, formula (2) is used. Therefore, as shown in step 24, the adjusted total effect value calculation module 14 calculates an adjusted total based on the expected total effect value, correlation coefficient, and R-squared value. Effect value. In the process of adjusting the total effect value calculation module 丨4 to calculate the total effect value, it is first determined whether the correlation coefficient is greater than spring zero. If the correlation coefficient is greater than zero, indicating that the evaluated person's performance is positive, the adjusted total effect value computing module 14 further determines how to adjust the total effect value according to whether the expected total effect value is greater than zero. If the expected total effect value is greater than zero', the adjusted total effect value calculation module 14 calculates the adjusted total effect value=the expected total effect value χ(1+R squared value); if the expected total effect value is less than zero, Then, the adjusted total effect value calculation module 14 calculates the adjusted total effect value=the expected total effect value X (1_R square value). On the other hand, if the adjusted total effect value calculation module 14 determines that the correlation coefficient is less than zero, indicating that the performance of the evaluated person is negative growth, the adjusted total effect value calculation module 14 calculates the adjusted total effect value = the expectation. The total effect value is parent|1+砰方値|. Therefore, in the case where the correlation coefficient 0.5 9 is greater than zero and the expected total effect value is greater than zero in the present embodiment, the adjusted total effect value module 14 calculates the adjusted total effect value = 46 outpatients. Then, the performance value calculation module 15 is based on the adjusted total effect value.
S 10 201232433 及一預定效果係數，運算該受評估人員之一績效值。在績 效值運算模組15運算績效值的過程中，是視調整後總效果 值之絕對值是否小於等於卜來決^績效值。若調整後總效 果值之絕對值小於等於丨，則該績效值等於零。 反之’若調整後總效果值之絕對值大於 π、=貝双 m :¾S 10 201232433 and a predetermined effect coefficient, the performance value of one of the evaluated persons is calculated. In the process of calculating the performance value by the performance value calculation module 15, it is determined whether the absolute value of the total effect value after adjustment is less than or equal to the performance value of the Bulaid. If the absolute value of the adjusted total effect value is less than or equal to 丨, the performance value is equal to zero. Otherwise ‘if the absolute value of the total effect value after adjustment is greater than π, = double double m : 3⁄4
算模組15進一步視調整後總效果值是否大於零來運算績效 值；若調整後總效果值大於零，則績效值運算模組15運算 該績效值=〇n|該調整後總效果値|)λ該預定效果係數，其中丨n為 自^然對數函數，·若調整後總效果值小於零，則績效值運算 杈組15運算該績效值= —(Ιη丨該調整後總效果値丨)八該預定效果係 數。其中，該預定效果係數是用以調整主政者對核發獎金 上下限的取捨態度。例如，在如纟】所示工作量為醫師每 1貫際Η診量的實施例中’該預定效果係數可經由實際數 據而模擬出一適當的數值範圍，例如介於3盘4之門 心=施例調整後總效果值為46次門診，;絕對值 杳，且其大於零，以及取狀效果係數〜的情況下， 算模組15運算㈣醫師之績效值為m。亦即 ”-可根據本發明自動化績效評估系統與方法所 算出的績效值丨丨〇，來發认歹鑿 運 _獎勵金。反之：:二°單位(如新台幣η。 所自動運/ 自動化績效評估系統與方法 Ρ方··Τ·效值為負值’表示該醫師的績效不，， …根據本發明自動化績效 Τ 的懲罰值。 ”口去相當於該負績效值 201232433 綜上所述，本發明自動化績效評估方法及系統運用了 學習曲線理論與迴歸分析，可將受評估人員的工作量數值The calculation module 15 further calculates the performance value according to whether the adjusted total effect value is greater than zero; if the adjusted total effect value is greater than zero, the performance value calculation module 15 calculates the performance value=〇n| the total effect after the adjustment 値| λ the predetermined effect coefficient, where 丨n is a self-correlation logarithm function, and if the adjusted total effect value is less than zero, the performance value operation group 15 calculates the performance value = - (Ιη丨 the total effect after the adjustment値丨) Eight of the predetermined effect coefficient. The predetermined effect coefficient is used to adjust the attitude of the ruling party to the upper and lower limits of the bonus issued. For example, in the embodiment where the workload shown by the doctor is a per-biological amount per physician, the predetermined effect coefficient can simulate an appropriate range of values via actual data, for example, between 3 discs = The total effect value after the adjustment of the example is 46 outpatients; if the absolute value is 杳, and it is greater than zero, and the effect coefficient ～ is taken, the calculation result of the module 15 is (4) the performance value of the physician is m. That is, "the performance value calculated by the automated performance evaluation system and method of the present invention can be used to recognize the 歹 _ _ rewards. Conversely:: two units (such as NT η. Automated transport / automation The performance evaluation system and method Ρ · 效 效 效 效 效 效 效 效 效 效 效 效 效 效 表示 表示 表示 ' ' ' ' ' ' ' ' 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该The automatic performance evaluation method and system of the present invention utilizes a learning curve theory and a regression analysis, and can calculate the workload value of the evaluator
自動轉換為績效值，藉以對！士 E 精對早成長以及陡成長者予以鼓勵 ’故碟實能達成本發明之目的。 惟以上所述者，僅為本發明之較佳實施例而已，當不 能以此限定本發明實施之範圍，即大凡依本發明巾請專利 範圍及發明說明内容所作夕鸽a 合所作之簡早的等效變化與修飾，皆仍 屬本發明專利涵蓋之範圍内。 【圖式簡單說明】 圖1是一統計圖表，說明-學習成長曲線，其中包含 一虛擬預測線； 統之 圖2疋一方塊圖，說明本發明自動化績效評估系 較佳實施例； ’' 方法之 圖3是-流程圖，說明本發明自動化績效評估 較佳實施例；以及 圖4是一統計圖表，說明本發明實施例中藉由將受評 估人員的實際工作量數據轉換成迴歸直線。Automatically convert to performance value, so you can! Shi E Jing encourages early growth and steep growth. The original dish can achieve the purpose of the present invention. However, the above is only a preferred embodiment of the present invention, and it is not intended to limit the scope of the practice of the present invention, that is, the simpleness of the invention according to the scope of the invention and the description of the invention. Equivalent variations and modifications are still within the scope of the invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a statistical diagram illustrating a learning growth curve including a virtual prediction line; FIG. 2 is a block diagram showing a preferred embodiment of the automated performance evaluation system of the present invention; ''method 3 is a flow chart illustrating a preferred embodiment of the automated performance evaluation of the present invention; and FIG. 4 is a statistical diagram illustrating the conversion of the actual workload data of the assessee into a regression line in an embodiment of the present invention.
S 12 201232433 【主要元件符號說明】 1 ..........自動化績效評估 13......... 系統 算模組 π.........實際工作量資料 14......... 庫 運算模組 12.........迴歸數據運算模 15......... 組 2 1〜2 5 — 期望總效果值運 調整後總效果值 績效值運算模組 步驟S 12 201232433 [Explanation of main component symbols] 1 ..........Automatic performance evaluation 13......... System calculation module π.........actual workload Data 14......... Library operation module 12.........Regression data operation mode 15.... Group 2 1~2 5 — Expected total effect value Adjusted total effect value performance value calculation module step
1313