A few days ago, I was experimenting with the neural network toolbox in MATLAB and suddenly felt like I was "squatting in the well"—like I had hit a wall. It wasn’t that the tool itself was difficult, but the data structure required for the neural network was a bit unusual. If you're not careful, it can easily lead to errors when using the toolbox. Below is the correct way to use the neural network toolbox, just for your reference.
Here’s a method that uses batch commands to call the Neural Network Toolbox:
```matlab
P = [0.1515 0.1501 0.1509 0.1504 0.1504 0.1500
0.1515 0.1501 0.1509 0.1504 0.1504 0.1500
0.1515 0.1501 0.1500 0.1504 0.1504 0.1500
0.1515 0.1500 0.1509 0.1504 0.1504 0.1500
0.1515 0.1500 0.1509 0.1504 0.1504 0.1500
0.1515 0.1500 0.1509 0.1504 0.1504 0.1500
0.9684 0.2792 0.877 0.7426 0.7228 0.2272
0.9838 0.2941 0.9181 0.7977 0.7702 0.2452
0.9922 0.3101 0.9475 0.8445 0.8227 0.2665
0.9953 0.3058 0.9625 0.8708 0.8637 0.2624
0.9982 0.3242 0.9797 0.9089 0.9001 0.3008
0.9995 0.3469 0.9917 0.9314 0.9282 0.3678
0.9998 0.3565 0.9948 0.9493 0.9525 0.4500];
T = [0.1521
0.6949
0.7064
0.7083
0.7560
0.7807
0.8182
0.8533
0.8677
0.8459
0.8910
0.9269
0.9496];
P = P'; T = T';
ff = newff(P, T, 13);
ff.trainParam.epochs = 15000;
ff = train(ff, P, T);
Y1 = sim(ff, P);
cf = newcf(P, T, 13);
cf.trainParam.epochs = 15000;
cf = train(cf, P, T);
Y2 = sim(cf, P);
plot(P, T, 'o-');
hold on;
plot(P, Y1, '^m-');
plot(P, Y2, '*-k');
title('newff & newcf');
legend('original data', 'newff result', 'newcf result', 0);
```
One important thing to note here is that `P` and `T` should be transposed first. In MATLAB, the neural network expects each column of the matrix to represent a training sample.
Now, if you want to use the Fuzzy Neural Network Toolbox, first install MATLAB. I won’t go into much detail about the installation process—just make sure you have it set up correctly.
Once you open MATLAB, type `anfisedit` at the command line, and the following interface will appear:
Click the "Load Data" button (in the first red box) to import your data. You can either load it from a file or from the workspace. What's important here is the format: the input part comes before the output. For example, if your system has 3 inputs and 1 output, your data should be structured as `[x1, x2, x3, y]`.
After importing the training data (selecting the "Training" data type), the toolbox will display the data accordingly:
Next, click on "FIS Properties" under the "Edit" menu to adjust the settings. A new window will pop up where you can define the number of inputs, outputs, and membership functions.
This is just the beginning. From here, you can fine-tune the fuzzy rules, optimize the model, and even export it for further analysis or deployment. The key takeaway is that proper data formatting and understanding the toolbox interface are crucial to avoid common pitfalls.
Capacitive touch film, also known as PCAP film or capacitive touch foil, is a transparent, flexible sensor layer that enables precise and responsive touch control. It uses projected capacitive (PCAP) technology to detect touch inputs through a glass or plastic surface, supporting multi-touch gestures with high accuracy and durability.
Â
These films are ideal for integration into interactive displays, smart windows, kiosks, digital signage, vending machines, and other smart surfaces. With their ultra-thin design, high light transmittance, and compatibility with curved or irregular surfaces, capacitive touch films offer design flexibility without compromising functionality.
Â
Key features include excellent sensitivity, support for up to 10-point multi-touch, anti-interference capability, and long-term stability. Available in various sizes and can be customized for different screen dimensions or industrial applications. Whether for retail, industrial control, automotive, or commercial use, capacitive touch films provide an innovative, seamless user interface solution.
PCAP film,capacitive touch foil,multi touch film,PACP touch foil,capacitive foil
Guangdong ZhiPing Touch Technology Co., Ltd. , https://www.zhipingtouch.com