Search

Selling and purchasing products online is made possible by e-commerce. For both service providers and clients, organizing and looking for items is a tedious procedure. The items must be organized and labeled, which takes up a lot of time. Product categorization is the process of automatically predicting a product's catalog route based on a predetermined catalog hierarchy in which all categories are formulated. Knowing how to add your goods to the most relevant category on any marketplace, including Flipkart, and Amazon is crucial to its selling. Categorization is a lengthy process that takes extensive study on the platform which has been improved with different methodologies used in this work. Machine Learning (ML) and Deep Learning (DL) models are used to sort items into recognized categories. Using information such as the item's title and summary, this model can properly classify it in each classification. Random Forest (RF) outperformed the other ML models, such as SVM, KNN, and Naive Bayes (NB), with an f1-score of 97 percent and a macro average of 94 percent. BERT model fared the best among the DL models (LSTM, CNN, BERT, and Hybrid CNN - LSTM model) with an f1-score of 97 percent and a macro average of 88 percent. [ABSTRACT FROM AUTHOR]

A malicious website, often known as a malicious URL, is a platform considering hosting unwanted content including spam, harmful advertisements, and dangerous websites. It is crucial towards quickly identify dangerous URLs. Blacklisting, regular expression, & signature matching techniques have all been employed in earlier investigations. These methods are utterly useless considering identifying new URLs, malicious URL variants, or URLs that have never been seen before. Machine learning-based solution that has been suggested can help towards solve this problem. Considering this kind about solution, in-depth study about feature engineering & feature representation about security artifact types, such as URLs, is necessary. Additionally, resources considering feature engineering & feature representation must be continuously improved towards support variations about current URLs or completely new URLs. Deep learning, machine learning, & artificial intelligence (AI) systems have recently been used towards achieve human-level performance in a number about areas & even surpass human eyesight in a number about computer vision applications. They can automatically extract best feature representation from raw inputs. We propose various algorithms, including SVM, Random forest, XgBoost, & AdaBoost, towards capitalise on & turn performance increase about them into cyber security area. [ABSTRACT FROM AUTHOR]

After a deep questioner among students, faculty members and higher education experts. The demand for qualified engineers in the specific field as understandably gone down. The situation is grimmer for tier-2 and tier-3 colleges in India. The purpose of opting engineering course is not to get expertise in the particular course, instead to get a job in government sector with valid degree. The problem identified in the field of engineering education towards improving the quality in education is addressed with the help of exploratory data analysis. The dataset used in our experiment is collect from UCI machine Learning Repository having 33 variables and 1044 observations. The contribution made in the paper is to identify the vital attributes using single and multi-variant regression techniques. To perform prediction using Decision Tree, Random Forest, Support vector Machine and compare their performance in terms of classification accuracy and F-Score. In addition to that a convolution Neural Network (CNN) model is established in which, the vital attributes identified using regression techniques are provided as inputs and weights at each stage is estimated using Gradient Decent algorithm with step size 0.5. The classification accuracy from 63% is improved to 97% with the help of CNN model in 26664 iterations. The finding of the present research states that student having backlogs are less frequently opting for higher studies. [ABSTRACT FROM AUTHOR]

The analysis of firewall logs is one of the most significant practices considered while monitoring network traffic to assess their impact. The log records of the Turkish Firat University's firewall device were analyzed using K-Nearest Neighbor (KNN), Random Forest (RF), and Deep Neural Network (DNN) classifiers. A comparison was conducted to measure the performance of the classifier in terms of accuracy, recall, precision, and F1 score. 65,532 records were examined using 12 attributes, where the action was identified as a label of these attributes because it handles the packets based on their features either allowing them to pass, blocking them, blocking their activity, or blocking the request itself. The result of this analysis indicated that the best algorithm that selects the best features according to appropriate action is Random Forest. [ABSTRACT FROM AUTHOR]

Agriculture is a field that plays an essential part in strengthening a country's economy, especially in agrarian countries like India, where agriculture and crop productivity play a large role in the economy. The research focuses on comparing machine learning and Deep learning algorithms in predicting total crop yield production. The parameters considered for the study are State name, District name, Year, Season, Crop, Area and Production. The dataset is resourced from the data.gov.in website. Random forest from Machine Learning and Sequential model from Deep learning are compared, and the performance metric considered for the study is R2 score. The objective is to assess how well the independent variable predicts the variance in the dependent variable. Random Forest algorithm achieved an R2 score of 0.89, whereas Deep Learning Sequential algorithm gave an R2 score of 0.29. [ABSTRACT FROM AUTHOR]

Life style disorders like diabetes, cardio vascular diseases (CVD) etc. are spreading like an epidemic to a pandemic. Heart disease (HD) besides other health consequences has serious social and economical implications for the individual and for the country. These conditions have multiple dimensions and can be controlled and prevented, if diagnosed and treated in time, by improving the overall personality/life style of an individual with the help of technology. It has been found from the literature that a huge number and combination of algorithms/classifiers with great performances/accuracies have already been employed to predict the heart disease (HD). SVM classifier has been predominantly deployed for classification. Neural networks (NNs) and ensembles have also been used which have also yielded high accuracy of over 95%. CNNs and deep learning techniques too have yielded significantly high accuracies of over 95% recently. It has already been established that ensemble techniques/models consisting of a combination of several related but different models like bagging, boosting, blending/stacking and Majority voting have performed better on (Cleveland) datasets with 91.6% accuracy than individual constituent models/classifiers-called base learners. Some ensemble algorithms/models like Random Forest algorithm have been found to have achieved 100% accuracy for heart disease prediction. Further, ensemble models solve the problems like over fitting etc. also and are more suitable for the imbalanced data sets like medical/heart datasets. Owing to the unique feature processing capabilities for processing big data which is heterogeneous, high dimensional, multivariate unstructured and unlabelled in nature, deep learning deals best with unstructured and unlabelled data i.e. unsupervised learning and do not need much data pre-processing in comparison with the data mining/machine learning and does not require the processes of filtering, normalization etc. that are manually done in case of machine learning. This review paper aims at analyzing the literature to review and compare different techniques utilized for HD prediction to identify the best base learner so that in future an effective deep learning based ensemble model could be proposed that can be applied for real world heart disease data sets with improved performance. [ABSTRACT FROM AUTHOR]

Cyber-attacks targeting organisations and individuals are getting worse day by day. With the technological improve-ments, the principles used to launch attacks are also varied. Intrusion Detection System (IDS) is a network solution to detect attacks and anomalies in the network. Leveraging Machine Learning for developing an IDS can make it more powerful. This work aims at evaluating the performance of machine learning models for Intrusion Detection System built based on CART Algorithm, ID3Decision tree algorithm, Random forest algorithm with Convolutional Neural Network. The model is built using KDD Cup99 Dataset. In this work we perform Linear Correlation and Mutual Information feature selection algorithms for ML algorithms and for Convolutional Neural Network algorithm being a deep learning technique feature selection happens automatically. Analysing various parameters including F1-score, ID3 has shown better results. [ABSTRACT FROM AUTHOR]

The process of building new dataset and the existence of such a data followed the urgent need for the existence of datasets that are specialized in educational lectures, so this will need an accurate classification algorithm to classify it, the benefit of classify such dataset is to minimize the workload of classifying each file manually and individually. In the present paper, authors perform experimentations for conducting an empirical deep learning study, especially, convolutional neural network, for three new datasets of educational lectures which are (PDF, Word and PowerPoint datasets), The three new datasets using real data educational resources lectures collected from various document projects of different universities and institutions. The architecture has been applied to the task of the text classification in the domain of the document with documents data-sets have been obtained from a variety of projects on actual document cases. The aim of the present study is to initially test the performance of each dataset (PDF, Word, and PowerPoint dataset) through using four machine learning classification algorithms which are (Bayes Net, Random Forest, Random Committee, and OneR). Second goal is experimenting the efficiency of the approach of the deep learning in the tasks of classification and after that, comparing the efficiencies with the efficiencies of traditional machine learning classification methods. Mainly two classification techniques used to maximize the benefits of the classification process, the first one is to use the deep learning algorithm which shows an accuracy of classifying file between (95 and 96%) for three new dataset files and standard machine learning algorithms (OneR, Random forest, Bayes net, and Random Committee) these algorithm shows accuracy 91% for PDF Dataset using random forest and random committee algorithms, for Word dataset the accuracy is 46% using random committee, and for the last dataset PowerPoint the accuracy is 77% using random forest, Therefore, we will choose Deep learning algorithm because it gives higher results and accuracy than machine learning algorithms. [ABSTRACT FROM AUTHOR]