PK IBM Warren McCulloch Walter Pits MP 1949 Hebb Hebb Hebb 145

Size: px

Start display at page:

Download "PK IBM Warren McCulloch Walter Pits MP 1949 Hebb Hebb Hebb 145"

兵旬姚
9 years ago
Views:

1 6

2 PK IBM Warren McCulloch Walter Pits MP 1949 Hebb Hebb Hebb 145

3 01 SQL Server 2008 Data Mining 1969 Seymour Papert Marvin Minsky XOR 1982 John Hopfield HNN 1986 David E. Rumelhart Back-propagation Papert Rosenblatt Perceptron Nerve Cells Soma Axon Dendrites Synapse 146

4 Excitatory Synapse Pulse Rate Inhibitory Synapse X1 X2 X3 W1 W2 W3 X1 W1 I j w ij O i j i I O w Activation Function Y1 147

5 01 SQL Server 2008 Data Mining S SQL Server 2008 Sigmoid Logistic Y=1/ (1+e -x ) 1 0 O j 1 1 e -I J SQL Server 2008 Back-propagation Network Input Layer Output Layer Hidden Layer 148

6 01 over-fitting Topology MAXIMUM_STATES n n Analysis Services 149

7 01 SQL Server 2008 Data Mining Error j O j (1 O j ) (T j O j ) O T Error j O j (1 O j ) Error k w jk ) k 6.1 Means Squared Error MSE w ij I* Error j O i w ij w ij w ij 150

8 01 06 l Learning Rate 0~1 The Gradient Steepest Descent Method 6.3 Step01 W 14 W 15 W 24 W 25 W 34 W 35 W 46 W Step (X1,X2,X3,Y)=(1,0,1,1) 4 I j = w ij O i j =1*0.2+0*0.4+1* =-0.7 i 1 1 O j = = = e -I j 1 e I j = w ij O i j =1*-0.3+0*0.1+1* =0.1 i 1 1 O j = = = e -I j 1 e -0.1 Step 6 I j = w ij O i j =0.332* * =-0.1 i 151

9 01 SQL Server 2008 Data Mining Step O j = = = e -I j Step 4 Error j O j (1 O j ) Error k w jk )=0.322( )(0.1311)(-0.3)= k 5 Step W 14 = ( )(1)=0.192 W 15 = ( )(1)= W 24 = ( )(0)=0.4 W 25 = ( )(0)=0.1 W 34 = ( )(1)= W 35 = ( )(1)=0.194 W 46 = (0.1311)(0.332)= W 56 = (0.1311)(0.525)= = ( )= = ( )= = (0.1311)= e 0.1 Error j O j (1 O j )(T j- O j )=0.474( )( )= Error j O j (1 O j ) Error k w jk )=0.525( )(0.1311)(-0.2)= k

10 /2 2 2 =4 2 CRIPS-DM Validation Set Holdout Set SQL Server 2008 HOLDOUT_PERCENTAGE 153

11 01 SQL Server 2008 Data Mining SQL Server

12 01 06 Co-linearity

13 01 SQL Server 2008 Data Mining BROADBAND CUST_ID int GENDER char(1) AGE int TENURE int CHANNEL char(1) AUTOPAY char(1) ARPB_3M float 156

14 06 01 CALL_PARTY_CNT float DAY_MOU float AFTERNOON_MOU float NIGHT_MOU float AVG_CALL_LENGTH float BROADBAND char(1) minute of usage MOU minute of usage MOU Case...when... DAY_MOU _RATIO: CASE WHEN (DAY_MOU+AFTERNOON_MOU+NIGHT_MOU)=0 THEN 1 ELSE DAY_MOU/(DAY_MOU+AFTERNOON_MOU+NIGHT_MOU) END 157

15 01 SQL Server 2008 Data Mining AFTERNOON_MOU MOU = MOU 6.6 Step01 Step Step BROADBAND Step Step Step06 Step SQL Server

16 01 SQL Server T-SQL WHERE

17 01 SQL Server 2008 Data Mining ~30.4 Tenure 38.3~ BROADBAND=1 2 BROADBAND= ~ ~

18 % % % % Lift Lift % 14.79% % Lift %

19 01 SQL Server 2008 Data Mining

20 Harrison, D. and Rubinfeld, D.L. `Hedonic prices and the demand for clean air', J. Environ. Economics & Management, vol.5, 81-1, 1978 HOUSE_ PRICE HOUSE_ID Int CRIME_RATE Float BUSINESS_RATIO Float RIVER_SIDE char(1) NOX Float ROOM_NUM Int HOUSE_AGE Float WIIGHTED_DISTANCE Float 163

21 01 SQL Server 2008 Data Mining HIGHWAY_INDEX Float TAX_RATE Float TEACHER_RATE Float BLACK_INDEX Float 1000(Bk )^2 LOW_STATUS_RATIO Float HOUSE_PRICE Float

22 LOW_STATUS_RATIO LOW_STATUS_RATIO WIIGHTED_DISTANCE TEACHER_RATE SQL Server 2008 HIDDEN_NODE_RATIO HOLDOUT_PERCENTAGE HIDDEN_NODE_RATIO * SQRT({ } * { })

23 01 SQL Server 2008 Data Mining HOLDOUT_SEED 0 MAXIMUM_INPUT_ ATTRIBUTES MAXIMUM_OUTPUT_ ATTRIBUTES MAXIMUM_STATES 100 SAMPLE_SIZE SAMPLE_SIZE * (1 - HOLDOUT_ PERCENTAGE/100) Pyramids 6.3 HIDDEN_NODE_RATIO HIDDEN_NODE_RATIO * SQRT({ } * { }) HIDDEN_NODE_RATIO 166

24 06 1 RIVER_SIDE 11 TEACHER_RATE RIVER_SIDE 1 0 TEACHER_RATE SQRT(1*14)=3.74 HIDDEN_NODE_RATIO *4= HIDDEN_NODE_RATIO *3=

第一章　緒論

第一章　緒論中華大學碩士論文題目 : 類神經網路與一般克利金法在空間內插之比較 Comparison of Artificial Neural Networks and Ordinary Kriging Method in Spatial Interpolation 系所別 : 土木工程學系碩士班學號姓名 :M0904044 薛