Online Transaction Fraud Detection Using Efficient Dimensionality Reduction and Machine Learning Techniques

Chenoori, Rathan Kumar; Kavuri, Radhika

doi:10.18280/ria.360415

Cited by 4 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The gradient-boosted trees approach is one of the most used and well-implemented in form of a decision tree. The XGBoost tree structure is shown in Figure 3 [35][36][37].…”

Section: Extreme Gradient Boost Classifiermentioning

confidence: 99%

A Comparative Analysis of Machine Learning Models for Crop Recommendation in India

Reddy,

Neerugatti

2023

RIA

View full text Add to dashboard Cite

Agriculture serves as the mainstay of India's economy, bearing a vital responsibility in nourishing an expanding populace. The thriving of this sector is contingent upon numerous variables, among which the choice of the optimal crop plays a pivotal role. The advent of Machine Learning (ML) has engendered a transformative impact on the agricultural sector by facilitating the prediction of suitable crops, contingent on soil attributes. This study undertakes the examination of diverse ML algorithms, encompassing Decision Tree, Linear Regression, Naï ve Bayes, Random Forest, Extreme Gradient Boosting (XGBoost), and Support Vector Machine, to assess their efficacy in recommending optimal crops based on soil parameters. The parameters under consideration include Phosphorus, Nitrogen, Potassium, Electrical Conductivity, pH, Organic Carbon, Boron, Iron, Zinc, Copper, Manganese, and Sulphur. The crop recommendations are focused on Rice, Cotton, and Jowar for the Kurnool district of Andhra Pradesh, India. Among the assessed models, it was observed that the XGBoost model surpassed others in terms of accuracy in determining the most suitable crop for the given soil parameters. The experimental findings substantiate the precision of the model in forecasting the apt crop, thereby underscoring the immense potential of ML in the agricultural domain. This investigation signifies a considerable stride towards optimal crop recommendation, thereby increasing the potential for enhanced yield and profitability. Through the incorporation of technological innovation, agriculture can be rendered more efficient, cost-effective, and sustainable, thus laying the groundwork for a promising future.

show abstract

“…The gradient-boosted trees approach is one of the most used and well-implemented in form of a decision tree. The XGBoost tree structure is shown in Figure 3 [35][36][37].…”

Section: Extreme Gradient Boost Classifiermentioning

confidence: 99%

A Comparative Analysis of Machine Learning Models for Crop Recommendation in India

Reddy,

Neerugatti

2023

RIA

View full text Add to dashboard Cite

show abstract

“…To cope with these matters, now that a number of educational institutions and researchers have begun to pay attention to the application of data-driven methods in education management, it's been found that students' learning needs can be predicted more accurately through the analysis of students' online learning behavior and feedback [9][10][11], and educational resources could be updated and optimized effectively via data analysis and management of knowledge libraries, so as to meet students' personalized learning needs [12][13][14][15][16]. However, existing studies mostly talk about every single analysis and prediction method, and few of them have been concerned about how to effectively combine these methods with knowledge library management.…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Prediction of Students’ Online Learning Needs and Optimization of Knowledge Library Management

2023

Int. J. Emerg. Technol. Learn.

View full text Add to dashboard Cite

Thanks to the advancement of information technology, online learning has become a crucial tool of modern education, and the management of modern education is facing the challenges of how to effectively predict students’ learning needs and how to optimize the management of the knowledge library to support these needs. However, existing data-driven prediction approaches are flawed in handling complex learning environments and timely adapting to changes, so this study attempts to solve these questions by exploring the correlation between learning needs, the extraction of knowledge linkages, and the optimization of knowledge libraries based on rule updates. In our work, a new method was proposed for extracting learning needs and knowledge linkages to more accurately identify and predict students’ learning needs, and a rule-based knowledge library management optimization method was introduced to allow the knowledge library to more flexibly adapt to students’ learning needs and the changes in educational resources. In this way, this study offers a comprehensive and flexible solution for education management via the combination of these two aspects, which not only increases student satisfaction and improves teaching quality but also reduces resource waste and gives students a more personalized and efficient learning experience. Furthermore, the methods and findings of this study could also be used as references for data-driven decision-making in other fields.

show abstract

“…A New Methodology of Knowledge Point Sequence Generation and Learning Path Recommendation by Knowledge Reasoning information overload, and learners always feel lost in the countless courses and knowledge points [5][6][7][8][9]. Besides, the conventional education mode is always rigescent and fixed, so it's impossible to adapt to the individual needs of learners and adjust the content according to their respective knowledge foundation [10][11][12][13][14][15].…”

mentioning

confidence: 99%

A New Methodology of Knowledge Point Sequence Generation and Learning Path Recommendation by Knowledge Reasoning

Zhou¹,

Liang²

2023

Int. J. Emerg. Technol. Learn.

View full text Add to dashboard Cite

Accurate knowledge point sequences and clear learning paths are the compass for learners to travel in the sea of massive knowledge, they can point out the way for learners and improve their learning experience. Generating knowledge point sequences and learning paths based on knowledge reasoning is conductive to optimizing the allocation of educational resources and improving the quality of higher education, and this has a profound influence on the reform of the entire educational field. In view of this, this study explored the generation of knowledge point sequences and the recommendation of learning paths based on knowledge reasoning. At first, the learning behavior of learners was subjected to collaborative analysis based on three aspects of knowledge points: learning frequency, learning duration, and pause/ skip frequency, and the specific method of generating subject knowledge point sequences based on the metrics of difficulty differences was given. Then, a sequence sampling method that matches the features of Entity-Relationship (ER) diagram was proposed, which enables the system to dynamically adjust the recommended knowledge points and learning paths according to learners’ learning progress with the help of biased random walks, thereby giving personalized and dynamic learning recommendations. At last, the validity of the proposed method was verified by experimental results.

show abstract

Online Transaction Fraud Detection Using Efficient Dimensionality Reduction and Machine Learning Techniques

Cited by 4 publications

References 22 publications

A Comparative Analysis of Machine Learning Models for Crop Recommendation in India

A Comparative Analysis of Machine Learning Models for Crop Recommendation in India

Data-Driven Prediction of Students’ Online Learning Needs and Optimization of Knowledge Library Management

A New Methodology of Knowledge Point Sequence Generation and Learning Path Recommendation by Knowledge Reasoning

Contact Info

Product

Resources

About