合肥生活安徽新聞合肥交通合肥房產(chǎn)生活服務合肥教育合肥招聘合肥旅游文化藝術(shù)合肥美食合肥地圖合肥社保合肥醫(yī)院企業(yè)服務合肥法律

        代寫159.740編程、代做c/c++,Python程序
        代寫159.740編程、代做c/c++,Python程序

        時間:2024-11-04  來源:合肥網(wǎng)hfw.cc  作者:hfw.cc 我要糾錯



        159.740 Intelligent Systems
        Assignment #2 
        N.H.Reyes 
        Letter Recognition using Deep Neural Nets with Softmax Units 
        Deadline: 4th of November 
        Instructions: 
        You are allowed to work in a group of 2 members for this assignment. 
        Your task is to write a program that implements and tests a multi-layer feed-forward network for 
        recognising characters defined in the UCI machine learning repository: 
        http://archive.ics.uci.edu/ml/datasets/Letter+Recognition
        Requirements: 
        1. Use QT to develop your Neural Network application. A short tutorial on QT, and a start-up 
        code that will help you get started quickly with the assignment is provided via Stream. 
        2. You may utilise/consult codes available in books and websites provided that you cite them 
        properly, explain the codes clearly, and incorporate them with the start-up codes provided. 
        3. Implement a multi-layer feed-forward network using backpropagation learning and test it on the 
        given problem domain using different network configurations and parameter settings. There 
        should be at least 2 hidden layers in your neural network. 
        h21 h11 X1
        X2
        F1
        F2 h12 h22
        OF1
        OF2
        δh21
        δh22 δh12
        δf1
        δf2
        δh11
        … … … … 
        X16
        Fm Hi Hj
        OFm
        Input node
        Legend: 
        hidden node
        output node = softmax unit
         Note that all nodes, except the input nodes have a bias node attached to it. 
        159.740 Intelligent Systems
        Assignment #2 
        N.H.Reyes 
        A. Inputs 
         16 primitive numerical attributes (statistical moments and edge counts) 
         The input values in the data set have been scaled to fit into a range of integer values 
        from 0 through 15. It is up to you if you want to normalise the inputs before feeding 
        them to your network. 
        B. Data sets 
         Use the data set downloadable from: 
         Training set: use the first 16,000 
         Test set/Validation set: use the remaining 4,000 
         Submit your training data, validation/test data in separate files. 
        C. Performance measure: 
         Mean Squared Error (MSE) 
         Percentage of Good Classification (PGC) 
         Confusion Matrix (only for the best Neural Network configuration found) 
        D. Training 
         Provide a facility for shuffling data before feeding it to the network during training 
         Provide a facility for continuing network training after loading weights from file (do not 
        reset the weights). 
         Provide a facility for training the network continuously until either the maximum 
        epochs have been reached, or the target percentage of good classification has been met. 
         For each training epoch, record the Mean Squared Error and the Percentage of Good 
        Classification in a text file. You need this to plot the results of training later, to 
        compare the effects of the parameter settings and the architecture of your network. 
        E. Testing the Network 
         Calculate the performance of the network on the Test set in terms of both the MSE and 
        PGC. 
        F. Network Architecture 
         It is up to you to determine the number of hidden layers and number of hidden nodes 
        per hidden layer in your network. The minimum number of hidden layers is 2. 
         Use softmax units at the output layer 
         Experiment with ReLU and tanh as the activation functions of your hidden units 
         Determine the weight-update formulas based on the activation functions used 
        4. Provide an interface in your program for testing the network using an input string consisting of 
        the 16 attributes. The results should indicate the character classification, and the 26 actual 
        numeric outputs of the network. (the start-up codes partly include this functionality already, for 
        a simple 3-layer network (1 hidden layer), but you need to modify it to make it work for the 
        multiple hidden layer architecture that you have designed). 
        5. Provide an interface in your program for: 
        A. Reading the entire data set 
        B. Initialising the network 
        C. Loading trained weights 
        D. Saving trained weights 
        E. Training the network up to a maximum number of epochs 
        159.740 Intelligent Systems
        Assignment #2 
        F. Testing the network on a specified test set (from a file) 
        G. Shuffling the training set. 
        6. Set the default settings of the user interface (e.g. learning rate, weights, etc.) to the best 
        configuration that delivered the best experiment results. 
        7. Use a fixed random seed number (123) so that any randomisation can be replicated empirically. 
        8. It is up to you to write the main program, and any classes or data structures that you may 
        require. 
        9. You may choose to use a momentum term or regularization term, as part of backpropagation 
        learning. Indicate in your documentation, if you are using this technique. 
        10. You need to modify the weight-update rules to reflect the correct derivatives of the activation 
        function used in your network architecture. 
        11. Provide graphs in Excel showing the network performance on training data and test data 
        (similar to the graphs discussed in the lecture). 
        12. Provide the specifications of your best trained network. Fill-up Excel workbook 
        (best_network_configuration.xlsx). 
        13. Provide a confusion matrix for the best NN classifier system found in your experiments. 
        14. Provide a short user guide for your program. 
        15. Fill-up the Excel file, named checklist.xlsx, to allow for accurate marking of your assignment. 
        Criteria for marking 
         Documentation – 30% 
        o Submit the trained weights of your best network (name it as best_weights.txt) 
        o Generate a graph of the performance of your best performing network (MSE vs. 
        Epochs) on the training set and test set. 
        o Generate a confusion matrix of your best network 
        o fill-up the Excel file, named checklist.xlsx
        o fill-up the Excel file, named best_network_configuration.xlsx
        o provide a short user guide for your program 
         System implementation – 70% 
        Nothing follows. 
        N.H.Reyes 

        請加QQ:99515681  郵箱:99515681@qq.com   WX:codinghelp





         

        掃一掃在手機打開當前頁
      1. 上一篇:DATA 2100代寫、代做Python語言編程
      2. 下一篇:ME5701程序代寫、代做Matlab設計編程
      3. ·代寫2530FNW、代做Python程序語言
      4. ·代寫CIS5200、代做Java/Python程序語言
      5. ·LCSCI4207代做、Java/Python程序代寫
      6. ·代寫COP3502、Python程序設計代做
      7. ·代做MLE 5217、代寫Python程序設計
      8. ·代寫ISAD1000、代做Java/Python程序設計
      9. ·代做COMP3811、C++/Python程序設計代寫
      10. ·代寫SCIE1000、代做Python程序設計
      11. ·代寫comp2022、代做c/c++,Python程序設計
      12. ·CVEN9612代寫、代做Java/Python程序設計
      13. 合肥生活資訊

        合肥圖文信息
        急尋熱仿真分析?代做熱仿真服務+熱設計優(yōu)化
        急尋熱仿真分析?代做熱仿真服務+熱設計優(yōu)化
        出評 開團工具
        出評 開團工具
        挖掘機濾芯提升發(fā)動機性能
        挖掘機濾芯提升發(fā)動機性能
        海信羅馬假日洗衣機亮相AWE  復古美學與現(xiàn)代科技完美結(jié)合
        海信羅馬假日洗衣機亮相AWE 復古美學與現(xiàn)代
        合肥機場巴士4號線
        合肥機場巴士4號線
        合肥機場巴士3號線
        合肥機場巴士3號線
        合肥機場巴士2號線
        合肥機場巴士2號線
        合肥機場巴士1號線
        合肥機場巴士1號線
      14. 短信驗證碼 酒店vi設計 NBA直播 幣安下載

        關(guān)于我們 | 打賞支持 | 廣告服務 | 聯(lián)系我們 | 網(wǎng)站地圖 | 免責聲明 | 幫助中心 | 友情鏈接 |

        Copyright © 2025 hfw.cc Inc. All Rights Reserved. 合肥網(wǎng) 版權(quán)所有
        ICP備06013414號-3 公安備 42010502001045

        主站蜘蛛池模板: 国产精品xxxx国产喷水亚洲国产精品无码久久一区 | 精品国产日韩亚洲一区| 色噜噜AV亚洲色一区二区| 亚洲精品日韩一区二区小说| 精品一区二区三区免费 | 麻豆一区二区三区蜜桃免费| 日韩精品一区二区亚洲AV观看| 国产丝袜无码一区二区视频| 亚洲精品精华液一区二区| 91精品一区二区综合在线| 亚洲国产一区二区三区青草影视| 日韩精品无码一区二区视频| 国产午夜精品一区二区三区小说 | 亚洲av无码一区二区三区观看| 国产人妖视频一区二区| 人妻少妇一区二区三区| 中文字幕一区二区人妻性色| 大伊香蕉精品一区视频在线| 国产乱码伦精品一区二区三区麻豆| 动漫精品第一区二区三区| 视频一区二区三区在线观看| 精品视频午夜一区二区| 一区二区三区精品高清视频免费在线播放| 欧美日韩综合一区二区三区| 福利国产微拍广场一区视频在线| 国产精品无码一区二区在线 | 亚洲一区二区三区AV无码| 亚洲一区二区三区AV无码| 日韩AV无码久久一区二区| 91精品福利一区二区| 亚洲综合无码一区二区痴汉| 天海翼一区二区三区高清视频| 大香伊蕉日本一区二区| 爆乳熟妇一区二区三区霸乳| 国产一区二区三区夜色| 日本中文字幕一区二区有码在线| 亚洲美女视频一区| 亚洲AV网一区二区三区| 亚洲国产精品一区二区九九| 中文字幕精品一区二区精品| 国产电影一区二区|