Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


E. Laxmi Lydia has a total of 152 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A Financial Time Series Forecasting Model Using Quasi-Recurrent Neural Networks and the Crown Porcupine Optimizer for Stock Market Risk Prediction

    by Ilyos Abdullayev, Elvir Akhmetshin, Emil Hajiyev, Zokir Mamadiyarov, Tatyana Khorolskaya, Laxmi Lydia

    Stock is a financial product described by flexible trading, high risk, and high return, which is favored by many investors. Investors may obtain sufficient returns by precisely approximating stock price developments. However, a stock price is affected by several factors, such as market conditions, macroeconomic situation, major economic and social events, and managerial decisions of companies. As a result, Stock Price Prediction (SPP) has consistently attracted attention and is an important field of investigation. Currently, Machine Learning (ML) is extensively utilized in SPP, but considerably more appropriate techniques are suggested for the prediction of stock prices, as numerous investigations have proved that Deep Learning (DL) has better efficacy than other methods. The paper proposes a Financial Time Series Forecasting Using Quasi Recurrent Neural Network and Crown Porcupine Optimizer (FTSF-QRNNCPO) method, offering an intelligent framework for financial stock market price prediction to accurately assess market volatility and forecast associated investment risks. The FTSF-QRNNCPO method begins with data preprocessing, comprising missing value handling, data cleaning, and normalization to prepare the input data effectively. The Secretary Bird Optimization Algorithm (SBOA) is employed for optimal Feature Selection (FS). Then, the Quasi-Recurrent Neural Network (QRNN) approach is employed for prediction. The Crown Porcupine Optimizer (CPO) approach is employed in the parameter tuning process. The FTSF-QRNNCPO technique was experimentally evaluated on a Tesla stock price dataset, demonstrating superior accuracy, achieving a significantly lower MAPE of 0.415% and outperforming previous models.


  • Open Access

    Semantic segmentation assisted deep ensemble feature learning model for skin-cancer detection and classification: SDENet

    by Ch. Srilakshmi, N. Ramakrishnaiah, E. Laxmi Lydia

    The last few years have witnessed rapid increase in skin cancer caused mortality rate. Despite innovations and growth in vision-computing and artificial intelligence technologies, the complex shapes, sizes, textural patterns and ambiguous edges limits the reliability of existing approaches. Nevertheless, unlike traditional approaches the deep learning methods have performed superior; yet, the demands for the superior skin-lesion segmentation, ROI-specific feature extraction and learning can’t be ruled out. Moreover, it requires addressing class-imbalance problems as well to avoid skewed learning and prediction. Considering it as motivation, in this paper a novel and robust semantic segmentation assisted deep ensemble feature learning environment for skin-cancer detection and classification (SDENet) is proposed. The proposed SDENet model is targeted to perform multi-class skin-cancer classification. To achieve it, the SDENet at first performs standard pre-processing followed by synthetic minority over-sampling (SMOTE) to alleviate class-imbalance problem. Subsequently, it performs firefly heuristic algorithm based Fuzzy C-means clustering to segment skin-lesions (say, ROI), which is followed by ROI-specific deep spatio-textural ensemble feature extraction and fusion (DeS-TEFF). Specifically, SDENet makes use of the AlexNet deep network, DenseNet121 and Gray level co-occurrence matrix (GLCM) feature extraction methods. Here, AlexNet serves high-dimensional information rich features, while DenseNet121 yields layer-wise learning and feature reuse driven feature-set. Performing horizontal concatenation over the AlexNet, DenseNet121 and GLCM features, the principal component analysis (PCA) feature selection was performed, which helped to avoid local minima and convergence. The selected features were normalized by means of the z-score normalization so as to avoid over-fitting problems. Finally, the normalized features were trained and classified by using Heterogenous Ensemble Classifier, embodying SVM, DT, Random Forest, Extra Tree Classifier and XGBoost classifiers. The maximum voting ensemble-based classification over HAM10000 dataset exhibited the average accuracy of 98.97%, precision 99.38%, recall 98.94% and F-Measure 0.99, confirming its superiority over other existing approaches for real-time skin cancer diagnosis purposes.


  • Open Access

    An Explainable AI-based Fraud Detection System Using Recursive Feature Elimination and Waterwheel Plant Optimization for Financial Transactions

    by Hafis Hajiyev, Emil Hajiyev, Mirzobek Avezov, Samariddin Makhmudov, Dilora Abdukhalikova, E. Laxmi Lydia

    Fraudulent transactions and the methods to detect them are an an important issue for financial organizations globally. The requirement for progressive fraud detection systems to protect properties and maintain customer trust is predominant for financial organizations, but particular factors make the development of efficient and effective fraud detection models a challenge. Deep Learning (DL) has greatly improved fraud detection accuracy by detecting intrinsic patterns, whereas interpretability techniques improve transparency and build trust by making predictions understandable to experts. This study presents a Fraud Detection System using Recursive Feature Elimination and Waterwheel Plant Optimization (FDS-RFEWPO) model for financial transactions. The aim is to perform a comprehensive evaluation of fraud detection in high-dimensional financial transactions using advanced techniques. Initially, the FDS-RFEWPO technique follows min-max-based data pre-processing to normalize the input data. For the feature selection process, the FDS-RFEWPO model employs the Recursive Feature Elimination (RFE) technique to select the most relevant features from the dataset. Furthermore, the Variational Autoencoder/Wasserstein Autoencoder (VAE/WAE) model is employed for fraud detection and classification. To further enhance model performance, the Waterwheel Plant Optimization (WPO) technique is employed for hyperparameter tuning, ensuring the selection of optimal parameters that contribute to improved accuracy. Finally, the Explainable Artificial Intelligence (XAI) technique applies Local Interpretable Model-Agnostic Explanations (LIME) to improve the transparency, interpretability, and trustworthiness of Artificial Intelligence (AI) methods by making their decision-making procedures clear to humans. To evaluate the performance of the FDS-RFEWPO model, a comprehensive experimental analysis is conducted using a financial fraud detection dataset. The comparison study of the FDS-RFEWPO model demonstrates a superior accuracy of 97.41% over existing techniques.


  • An Intelligent Decision Support Systems for Financial Fraud Detection Using Pythagorean Neutrosophic Bonferroni Mean Approach with Machine Learning Models

    by E. E., Ranokhon Sharofutdinova, Aleksey Ilyin, K. Shankar, E. Laxmi Lydia, Umidjon Matyakubov

    Neutrosophy has developed as a generalization to fuzzy logic and is being employed in the research field in many areas such as set theory, logic, and others. Neutrosophic Logic is one of the neonate study regions and its intention is assessed to have the percentage of truth in a subset T, the percentage of falsity in a subset F, and the percentage of indeterminacy in a subset I. Recently, financial fraud has become a highly major issue, which results in severe consequences across firm sectors and affects people’s everyday lives. Therefore, financial fraud recognition is critical for the prevention of the regularly overwhelming effects of financial fraud. It includes differentiating fraudulent financial data from accurate data and permitting decision-makers to progress suitable plans to reduce the effect of fraud. Over the past few years, Artificial intelligence (AI), mainly machine learning (ML) systems, turned out to be the highest thriving model in fraud detection. This study presents a novel Intelligent Decision Support System for Financial Fraud Detection Using Pythagorean Neutrosophic Bonferroni Mean (IDSSFFD-PNBM) model. The main intention of the IDSSFFD-PNBM algorithm is to enrich the detection model for financial fraud using advanced optimization models. Initially, the z-score normalization is applied in the data normalization stage for converting input data into a beneficial format. Besides, the proposed IDSSFFD-PNBM designs a grasshopper optimization algorithm (GOA) for the selection of feature processes to enhance the efficiency and performance of the model. For the detection and classification procedure, the pythagorean neutrosophic bonferroni mean (PNBM) model has been employed. Additionally, the firefly optimization algorithm (FFOA)-based hyperparameter range method has been done to heighten the recognition outcomes of the PNBM system. The experimental evaluation of the IDSSFFD-PNBM technique takes place using a benchmark dataset. The experimental results indicated an enhanced performance of the IDSSFFD-PNBM technique compared to recent approaches


  • Harnessing Single-Valued Linguistic Complex Neutrosophic Set based Arabic Sentiment Classification on Natural Language Processing

    by Aigul Aigul, Alfiya Yarullina, Leysan Akhmetova, Barno Shamuratova, E. Laxmi Lydia

    Neutrosophic logic (NL) goes further by introducing a third component: indeterminacy. Each logical proposition in NL belongs to three degrees: truth (T), indeterminacy (I), and false (F), each taking value within the range of zero and one. This allows the processing and representation of uncertain, incomplete, and inconsistent data in a superior way. NL finds it beneficial in partially contradictory, partially known, and partially unknown scenarios, it becomes an effective instrument for applications in fields such as information fusion, artificial intelligence, and data analysis, where logical framework might be unsuccessful in handling the nuances and complexities of real-time data. Recently, Arabic sentiment analysis has become a hot research topic, which mainly intends to recognize sentiments that exist in Arabic social media. Therefore, this study introduces a Single-Valued Linguistic Complex Neutrosophic Set based Arabic Sentiment Classification (SVLCNS-ASC) technique on NLP applications. The presented SVLCNS-ASC technique undergoes Arabic data pre-processing and Glove word embedding process. For sentiment recognition, the SVLCNS-ASC technique applies the SVLCNS model, which enables to identification of various kinds of sentiments. At last, the performance of the SVLCNS model can be boosted by the use of artificial bee colony (ABC) based parameter-tuning approach. The results of the SVLCNS-ASC system has been studied on Arabic database. The experimental values indicate the supremacy of the SVLCNS-ASC approach compared to recent models.

  • Exploring Intuitionistic Fuzzy-Valued Neutrosophic Multiset Technique for High-Dimensional Financial Data Classification in Complex Systems

    by Hafis Hafis, Emil Hajiyev, Zarnigor Ilkhamova, Elena Klochko, E. Laxmi Lydia

    In decision-making, neutrosophic set allows for the information representation with three membership functions: truth (T), indeterminacy (I), and false (F). Each component in a neutrosophic set has membership, non-membership, and indeterminacy degrees that are independent and range from 0 to 1. This makes neutrosophic set especially suitable in complex decision-making scenarios where information is contradictory, incomplete, or ambiguous, which enables robust and more nuanced analysis and solutions. A large portion of finance companies experience problems handling vast amounts of data. These data are often left unstructured and unorganized. Therefore, it is necessary to classify them to exploit it. Data classification also simplifies to use, locating, and retrieval of information. It becomes vital while handling risk management, legal discovery, data security, and compliance. Therefore, this manuscript presents an Intuitionistic Fuzzy-Valued Neutrosophic Multiset based Financial Data Classification (IFVNMS-FDC) technique in Complex Systems. The main aim of the IFVNMS-FDC technique is to recognize and categorize the financial data into respective classes. To do so, the IFVNMS-FDC technique initially uses min-max scalar as a pre-processing step. Besides, the high-dimensional financial data can be handled by the design of whale optimization algorithm (WOA) based feature selection. Finally, the IFVNMS-FDC technique derives IFVNMS technique for the identification of various classes related to the financial data. A wide-ranging experiments were involved in exhibiting the performance of the IFVNMS-FDC technique. The experimental values depicted that the IFVNMS-FDC method obtains reasonable performance on financial data recognition.