https://github.com/manoharvit/digits-recognition-system-using-machinelearning
The approach and dataset determines the system's accuracy and efficiency. For handwritten digit recognition, this paper provides a reasonable overview of machine and deep learning techniques such as SVM (Support Vector Machine), CNN (Convolutional Neural Network), and KNN (K Nearest Neighbor). It also shows which algorithm is the most effective at doing digit recognition. It also provides a comparison among different algorithms based on their accuracy, so that the most accurate method with the minimum errors can be used in distinct handwritten digit recognition applications.
https://github.com/manoharvit/digits-recognition-system-using-machinelearning
Last synced: 5 months ago
JSON representation
The approach and dataset determines the system's accuracy and efficiency. For handwritten digit recognition, this paper provides a reasonable overview of machine and deep learning techniques such as SVM (Support Vector Machine), CNN (Convolutional Neural Network), and KNN (K Nearest Neighbor). It also shows which algorithm is the most effective at doing digit recognition. It also provides a comparison among different algorithms based on their accuracy, so that the most accurate method with the minimum errors can be used in distinct handwritten digit recognition applications.
- Host: GitHub
- URL: https://github.com/manoharvit/digits-recognition-system-using-machinelearning
- Owner: ManoharVit
- Created: 2022-01-20T09:16:13.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2023-06-28T16:36:19.000Z (about 2 years ago)
- Last Synced: 2024-12-30T04:48:54.278Z (7 months ago)
- Language: Jupyter Notebook
- Size: 116 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Digits-Recognition-System-using-MachineLearning
In this current age of digitization, handwriting recognition plays a vital role in information processing. A lot of information is available in paper format, and digital processing files are cheaper than processing written paper files. Digit recognition projects can be helpful for professional and commercial applications and will have practical applications in our daily lives, and can be of great help to the visually weaken people. With the help of this project, in the future we can quickly process data, like bank statements, forms, emails and more.
The design of such a system necessitates the creation of a machine that can recognise and classify images of handwritten numbers as 10 digits (0–9). Handwritten digits from the MNIST database have been known in the community for decades as a way to reduce error rates using various classifiers and settings. If the system reads a digit wrongly, it can cause significant damage, which is not desirable. As a result, in these real-world applications, a high-accuracy algorithm is necessary.
The approach and dataset determines the system's accuracy and efficiency. For handwritten digit recognition, this paper provides a reasonable overview of machine and deep learning techniques such as SVM (Support Vector Machine), CNN (Convolutional Neural Network), and KNN (K Nearest Neighbor). It also shows which algorithm is the most effective at doing digit recognition. It also provides a comparison among different algorithms based on their accuracy, so that the most accurate method with the minimum errors can be used in distinct handwritten digit recognition application
## Block diagram of system architecture of this project
## RESULTS AND DISCUSSIONS
After building all three algorithms, SVM, KNN, and CNN, we compared their accuracies and execution times to gain a better understanding. All of the models listed above have had their best accuracy taken into account. After testing all of the models, we discovered that SVM has the most accuracy on training data, whereas CNN has the highest accuracy on testing data. We also compared the execution times to acquire a better understanding of how the algorithms work. In general, an algorithm's execution time is proportional to the number of operations it has completed. So, to acquire the best result, we trained our deep learning model for 16 epochs and SVM models according to norms. SVM required the least amount of time to operate, whereas CNN required the most time.