How can you increase the accuracy of a SVM classifier?
Answers from top 6 papers
More filters
| Papers (6) | Insight |
|---|---|
865 Citations | Our results show that the SVM achieves a higher level of classification accuracy than either the ML or the ANN classifier, and that the SVM can be used with small training datasets and high‐dimensional data. |
11 Citations | Our experiments show that the SVM classifier provides a better performance by applying our compressing and balancing approach. |
11 Jun 2017 41 Citations | The experimental results show that the offered approach allows increasing the classification quality of the SVM classifier. |
01 Jan 2016 30 Citations | Results show that the SVM classifier achieves the highest accuracy rate, with 96.06% compared with other classifiers. |
| Experimental results show that the new classifier has good accuracy compared with the classic SVM, while the training is significantly faster than several other SVM classifiers. | |
| The experimental results show that the proposed method has a competitive learning ability and acquires better accuracy than SVM. |
Related Questions
How to improve the search accuracy?5 answersTo improve search accuracy, several methods have been proposed in the literature. One approach involves classifying HTML tags and assigning weighting coefficients to different types of tags, which helps in calculating the matching relevance of a search word and an HTML page. Another method focuses on improving graph search accuracy by establishing selection standards for the layout, removing short edges, and searching for specific graphs meeting size requirements. Additionally, the Best Order Score Search algorithm has been developed, which optimizes the model score by traversing permutations in a different way, resulting in more accurate results for larger and denser graphs. Furthermore, a method has been devised to improve search precision by splitting acquired words, associating them with preset classes, and combining adjacent words into a single word based on their associated class. These approaches aim to enhance search accuracy and provide more accurate and relevant results.
How can we improve the accuracy of time series prediction models?5 answersTo improve the accuracy of time series prediction models, several approaches have been proposed in the literature. One approach is to use deep learning techniques, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to extract complex temporal features from the time series data. Another approach is to incorporate attention mechanisms into the forecasting models, which can enhance the model's ability to capture important temporal patterns. Additionally, the use of generative adversarial networks (GANs) has been shown to improve the forecasting capability by reducing error accumulations and considering interactions between past data and future covariates. Furthermore, the augmentation of small sample datasets using sliding window techniques can help improve the accuracy of model selection. Overall, these approaches have demonstrated their effectiveness in improving the accuracy of time series prediction models.
How can we improve the accuracy of speech recognition systems?5 answersImproving the accuracy of speech recognition systems can be achieved through several methods. One approach is to use multiple automatic speech recognition (ASR) engines and generate multiple transcriptions for a given audio input. These transcriptions can then be combined and selected by a meta-speech engine to generate a more accurate response. Another method involves the use of geographic acoustic models (Geo-AM) and geo-specific language models (Geo-LMs) to address challenges such as multi-dialect and long-tailed distribution of points of interest (POI) names. The Geo-AM utilizes dialect-specific input features and top layers, while the Geo-LMs improve recognition accuracy for long-tailed and homophone POI names by selecting specific language models based on the user's geographic location. Additionally, creating an exclusive speech database for individual users and continuously updating it with correctly recognized speech data can enhance recognition accuracy, particularly for high-frequency users. Finally, the use of multi-head self-attention based attention methods, such as head fusion, can improve the recognition accuracy of speech emotion recognition (SER) models by generating attention maps with multiple attention points.
How can we improve the accuracy of fraud detection systems?4 answersTo improve the accuracy of fraud detection systems, researchers have proposed various approaches. One approach is the use of machine learning models, which have shown to be more accurate compared to rule-based methods. Additionally, reducing the cardinality of variables in a dataset of fraud transactions has been found to improve the classification performance. Another method involves utilizing the multi-view labeled and unlabeled data for fraud detection through a semi-supervised attentive graph neural network. One-class classification methods, such as SVDD, have also been applied to improve fraud detection by not requiring fraud samples for training. Furthermore, an extension of DBSCAN called REDBSCAN has been developed to reduce the number of samples while preserving the shape of the data, resulting in improved performance and speed of the fraud detection process.
How does the amount of noise in the data affect the accuracy of SVMs?4 answersThe amount of noise in the data has a significant impact on the accuracy of Support Vector Machines (SVMs). SVMs are known to be sensitive towards noise, which can affect their generalization ability and stability with respect to re-sampling. However, there are approaches that have been proposed to handle noise sensitivity in SVMs. For example, a modified Pin-SVM model has been developed to appropriately trade-off generalization ability, sparsity, and noise insensitivity. Additionally, a new adaptive memetic algorithm called PCA2MA has been proposed to select valuable training data for SVMs, which helps improve classifier score and speed up the classification process by reducing the number of support vectors. These approaches demonstrate the importance of addressing noise in the data to enhance the accuracy of SVMs.
How to improve the accuracy of SVM?6 answers