% Train a two layer neural network with the Levenberg-Marquardt % method. % % If desired, it is possible to use regularization by % weight decay. Also pruned (ie. not fully connected) networks can % be trained. % % Given a set of corresponding input-output pairs and an initial % network, % [W1,W2,critvec,iteration,lambda]=marq(NetDef,W1,W2,PHI,Y,trparms) % trains the network with the Levenberg-Marquardt method. % % The activation functions can be either linear or tanh. The % network architecture is defined by the matrix NetDef which % has two rows. The first row specifies the hidden layer and the % second row specifies the output layer.
標簽: Levenberg-Marquardt desired network neural
上傳時間: 2016-12-27
上傳用戶:jcljkh
This function calculates Akaike s final prediction error % estimate of the average generalization error for network % models generated by NNARX, NNOE, NNARMAX1+2, or their recursive % counterparts. % % [FPE,deff,varest,H] = nnfpe(method,NetDef,W1,W2,U,Y,NN,trparms,skip,Chat) % produces the final prediction error estimate (fpe), the effective number % of weights in the network if it has been trained with weight decay, % an estimate of the noise variance, and the Gauss-Newton Hessian. %
標簽: generalization calculates prediction function
上傳時間: 2016-12-27
上傳用戶:腳趾頭
MFC 視窗程式設計,視窗作業環境經多年試鍊,視窗應用程式於架構上已然出現了明顯的分類; 即便是架構不同,其間也存在著諸多共同點,例如:它們通常的都需要有功能表、 工具列等控制元件的設計,需要有用來動態管理資料的矩陣(arrays)、表列(lists) 等物件類別。
上傳時間: 2016-12-30
上傳用戶:lixinxiang
【歐拉算法】 微分方程的本質特征是方程中含有導數項,數值解法的第一步就是...歐拉(Euler)算法是數值求解中最基本、最簡單的方法,但其求解精度較低,一般不在...對于常微分方程: dy/dx=f(x,y),x∈[a,b] y(a)=y0 可以將區
上傳時間: 2014-01-09
上傳用戶:www240697738
以單輸入單輸出探討輸出可變結構控制的使用,對於使用者可以更能了解控制響應
標簽: 控制
上傳時間: 2014-01-13
上傳用戶:wweqas
#include "iostream.h" #include "iomanip.h" #define N 20 //學習樣本個數 #define IN 1 //輸入層神經元數目 #define HN 8 //隱層神經元數目 #define ON 1 //輸出層神經元數目 double P[IN] //單個樣本輸入數據 double T[ON] //單個樣本教師數據 double W[HN][IN] //輸入層至隱層權值 double V[ON][HN] //隱層至輸出層權值 double X[HN] //隱層的輸入 double Y[ON] //輸出層的輸入 double H[HN] //隱層的輸出
標簽: define include iostream iomanip
上傳時間: 2014-01-01
上傳用戶:凌云御清風
三維曲線曲面比較演示系統程序設計 設計一個圖形用戶界面(GUI)演示常見的三維函數圖形,至少包含“三維繪圖” 、“選項” 、“退出”等菜單,三維繪圖的包括:參數方程x=e-t/20cos(t), y= e-t/20sin(t),z=t其中t 為0到2π、參數方程x=t,y=t2,z=t3其中t為0到1之間(在同一圖形界面中分別繪制它們的三維曲面和三維曲線圖)。“選項”菜單主要包括:網格開關,圖例開關,坐標邊框開關,色度空間選擇菜單,曲線顏色菜單。
上傳時間: 2017-01-10
上傳用戶:hasan2015
Overview Input Clock = 24Mhz Preview VGA 15fps @ 60Hz VGA 12.5fps @ 50Hz Capture VGA 15fps @ 60Hz VGA 12.5fps @ 50Hz Output Format YCbCr 4:2:2 (ITU 656) YCbCr to RGB conversion R = Y + (351*(Cr – 128)) >> 8 G = Y – (179*(Cr – 128) + 86*(Cb – 128))>>8 B = Y + (443*(Cb – 128)) >> 8
上傳時間: 2013-12-24
上傳用戶:遠遠ssad
學位論文;運動物體跟蹤方法主要包括卡爾曼濾波,Mean-shift,Camshifi算法,粒子濾波器,Snake模型等;應用卡爾曼濾波方法設計了一套煤礦礦工出入自動監測系統;提出了一種新的基于高斯混合模型的顏色特征提取方法,該方法克服了現有的Camshift算法Continuousl y Adaptive eanshift中跟蹤目標特征提取精確度低和計算復雜度高的缺陷
上傳時間: 2017-01-14
上傳用戶:JasonC
一個基于GTK+的單詞數值計算器,1、 按照規則計算單詞的值,如果 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z 26個字母(全部用大寫)的值分別為 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26,如: WINJACK這個單詞的值就為:W+I+N+J+A+C+K=23+9+14+1+3+11=71% HARDWORK=H+A+R+D+W+O+R+D=8+1+18+4+23+15+18+11=98% LOVE=L+O+V+E=12+15+22+5=54% LUCK=L+U+C+K=12+21+3+11=47% ATTITUDE= A+T+T+I+T+U+D+E=1+20+20+9+20+24+4+5=100% 2、對程序的界面布局參考如下圖所示,在第一個單行文本框輸入一個單詞,點擊“計算”按鈕,按照以上算法計算出該單詞的值。 3、如果在最下面的單行文本框輸入一個文件路徑,此文件每行記錄一個單詞,那么經過程序計算出各個單詞的值,并把結果輸出到當前目錄下result.txt文件中。如果文件不存在,應該提示錯誤。
上傳時間: 2014-01-11
上傳用戶:康郎