Leveraging Dynamic ANN for Early Detection of Financial Distress | Chapter 2 | Mathematics and Computer Science: Contemporary Developments Vol. 10

Microfinance Institutions (MFIs) play a crucial role in Kenya, contributing approximately 7% to the economy according to the Central Bank of Kenya. Despite their success in promoting financial inclusion, MFIs face challenges such as poor management, competition, and unfavorable business conditions, which can lead to financial distress. In Kenya's economic landscape, financial hardship has been a growing concern since 2019 after the onset of Covid 19, leading to the closure of organizations as they are unable to meet their financial obligations and expectations. This study presents a financial distress prediction dynamic model using Artificial Neural Networks (ANN) using financial ratios as input features where each node represents a single financial metric utilizing their power to measure financial health, stability and standardization across periods of organizations. Data was normalized using a min-max scaling approach, setting values between 0 and 1 for better model convergence and split to training and testing set in a ratio 8:2. The perceived dynamic ANN model utilizes the Multilayer Perceptron Networks (MLP), which is modeled to be adaptable, scalable and adjustable with the ability to self-update its architecture and parameters achieving a 94% accuracy, with strong recall and precision of 92% showing predicted financially distressed instances that were actually distressed. Further, the model demonstrated a ROC-AUC score of 0.99, demonstrating its effectiveness in distinguishing between distressed and non-distressed instances. The model's balanced F1 score of 87% further highlights its value as an Early Warning System (EWS) for financial management, helping organizations make informed decisions early enough to avoid financial crises. To further improve the EWS, it is recommended to integrate a dynamic adjustment mechanism within the ANNGARCH framework. This enhancement aims to capture volatility in predictions and refine accuracy for risk assessment, which will be explored in future research.

Author (s) Details

Abel M. Nyabera

Department of Mathematics, Multimedia University of Kenya, Kenya.

 

Cynthia Ikamari
Department of Finance and Accounting, Multimedia University of Kenya, Kenya.

 

George Mocheche
Department of Mathematics, Multimedia University of Kenya, Kenya.

 

Please see the book here:- https://doi.org/10.9734/bpi/mcscd/v10/3252  

Predicting Rice Crop Yields in Prayagraj: A Comparison of Regression Techniques and Neural Networks | Chapter 3 | Geography, Earth Science and Environment: Research Highlights Vol. 5

This study investigates rice crop yield prediction in Prayagraj District, Uttar Pradesh, using twenty-nine years of historical weather and crop yield data from 1991 to 2019. The analysis divided the dataset into a calibration segment comprising 90% of the data over 26 years, with the remaining 10% used for validation. In our approach, 75.9% of the data was used to train an artificial neural network (ANN) model, and 24.1% was employed for testing and validation to ensure 100% model efficiency evaluation. Techniques such as stepwise linear regression and neural networks were applied to predict rice yields. The performance of these models was measured primarily by the Normalized Root Mean Squared Error (nRMSE), with the regression-based model achieving superior performance, indicated by the lowest nRMSE values. The study also noted the critical role of Bright Sunshine Hours, which demonstrated a significant predictive power with an nRMSE of 0.00025 and a coefficient of determination of 0.94.

 

Author (s) Details

 

Nilesh Kumar Singh
Sam Higginbottom University of Agriculture Technology and Sciences, Allahabad, Uttar Pradesh, India.

 

Shraddha Rawat
Sam Higginbottom University of Agriculture Technology and Sciences, Allahabad, Uttar Pradesh, India.

 

Please see the book here:- https://doi.org/10.9734/bpi/geserh/v5/4249

Designing of a Self-Learning Artificial Neural Network Controller for Critical Heating, Ventilation and Air Conditioning Systems | Chapter 2 | Advances and Challenges in Science and Technology Vol. 9

 Artificial affecting animate nerve organs networks (ANN) has emerged as a effective learning method to perform complex tasks in very non linear dynamic atmospheres. This work addresses the stability and adeptness problems accompanying standard Heating, Ventilation and Air-Conditioning (HVAC) systems by executing a self-learning ANN boss. Although traditional plans such as Proportional, Integral, and Derivative (PID), On-Off controllers, thus are used, they abandon to give intelligence and encourage mathematical complicatedness in implementation. They take more protracted to acquire a large size of stability, use plenty energy, and produce oscillations and peak overshoots. This work focuses on engaging a self-learning ANN located intelligent boss to scheme the air cooling whole. It uses the user's inputs to estimate the fan and water flow speed so that provide comfort accompanying the least amount of strength consumption and relieving time. The type of practice checked here is appropriate to many other types of non-uninterrupted control issues.  In neural topology, the Back Propagation (BP) method has been secondhand. The PID and Self Learning ANN controllers have been distinguished, and MATLAB Simulink has been used to display the results. Using the Self Learning ANN boss architecture, honest-time fittings for the HVAC system has happened created and distinguished with the PID controller.

Author(s) Details:

Raghavan Chandran Ilambirai,
Department of EEE, SRMIST, Kattankulathur, Tamil Nadu, India.

Shanmugapriya Subramaniyan,
Department of EEE, SRMIST, Kattankulathur, Tamil Nadu, India.

Geethanjali Subramaniyan,
Department of EEE, SRMIST, Kattankulathur, Tamil Nadu, India.

Please see the link here: https://stm.bookpi.org/ACST-V9/article/view/12618

Neural Network Approaches to Simulation of Spring Flood in Sites of the Lena River| Chapter 1 | New Frontiers in Physical Science Research Vol. 6

 The Republic of Sakha (Yakutia), accompanying an extent of 3103.2 thousand km2, is one of the best constituent bodies of the Russian Federation accompanying a piercingly continental environment and a large group of abundant and narrow rivers and streams. The Lena River is individual of the most protracted waterways of Russia, flows through all republic from cold to northward through the region accompanying different of or in the atmosphere zones, this leads to a a lot of compressing experiences during the spring floods accompanying a big relating to space and material range.Spring-summer floods cause mammoth damage to the domain, inducing inundation of vast domains and objects of the civil saving, that determines the pertinence and need to cultivate and correct flood guessing methods to implement convenient measures for fear that and humble the risk of inundation, especially in laboriously inhabited extents and districts with complex foundation, in the way that strategically main conveniences, underwater crossings of main pipelines, bridges and capacity lines. The projected forms manage possible to estimate water levels all the while spring floods two together on the action of calling a time succession from prior principles and reversion dependences on miscellaneous determinants (iceberg girth, temperature, etc.) and accompanying enough veracity, that is shown for one results of guessing maximum water levels e.g. of two portions of the Lena River, based on classical dossier.

Study about Statistical Normalization and Back Propagation | Chapter 04 | Novel Research Aspects in Mathematical and Computer Science Vol. 2

 Artificial Neural Networks (ANNs) are a prominent machine learning technique that has lately been used in a variety of medical and medically related industries. Diabetes is a chronic condition that develops when blood sugar levels are too high. Early diagnosis of the condition may help to reduce consequences. The Artificial Neural Network is a well-known nonlinear data classifier. When building the network model, there are a few main aspects to consider. The input vectors' normalisation, learning rate parameter, and network structure. The proposed study examines several normalising approaches used in back propagation neural networks to improve the trained network's dependability. Based on the experimental results, the performance of the classifier model may be improved.

Author(S) Details

T. Jayalakshmi
LRG Government Arts College For Women, Tiruppur, India.

A. Santhakumaran
LRG Government Arts College For Women, Tiruppur, India.

View Book:- https://stm.bookpi.org/NRAMCS-V2/article/view/6790

Assessment of Artificial Neural Network-based Induction Motor Fault Classifier Using Continuous Wavelet Transform| Chapter 12 | Novel Perspectives of Engineering Research Vol.7

 Due to their great benefits over other types of electric motors, induction motors are widely employed in industrial, commercial, and household applications. These motors are subjected to a wide range of operating stresses, which can lead to failure. The most common repeated failures in induction motors are bearing faults, stator interturn faults, and fractured rotor bars. For reliable and cost-effective operation, early diagnosis of induction motor defects is crucial. Induction motor faults and failures can cause long periods of downtime and significant maintenance and revenue losses. The cost of purchasing and installing equipment is often less than half of the entire cost of maintenance over the machine's lifetime. Maintenance costs range from 15% to 40% of the overall cost, and they might reach as high as 80% of the whole cost. The failure of a heavily loaded equipment can often bring an entire industry process to a standstill. The demand for automated manufacturing systems with effective monitoring and control capabilities has grown in response to the growing demand for high-quality, low-cost production.

An induction motor's condition monitoring and fault diagnosis are crucial in the production process. By enabling for the early detection of catastrophic failures, it can reduce maintenance costs and the danger of unanticipated breakdowns. Vibration monitoring, heat monitoring, chemical monitoring, and acoustic emission monitoring are just a few of the condition monitoring methods available, but they all require expensive sensors or specialised instruments. Current monitoring, on the other hand, does not necessitate the purchase of additional expensive sensors because basic electrical quantities such as voltage and current are easily measured by voltage and current transformers, which are always provided as part of the protective system. As a result, current monitoring is non-intrusive and can be used even if the motor is located far away from the control centre. As a result, MCSA demonstrates that it is a low-cost online nondestructive fault diagnostic and detection system that can accurately identify motor defects.

The findings of experiments using signal processing and artificial neural networks to identify numerous faults in induction motors are presented in this chapter. Motor line currents recorded under various fault circumstances were analysed using the continuous wavelet transform. For fault characterisation, a feedforward neural network was employed with fault features retrieved using the continuous wavelet transform.

Author(s) Details:

A. U. Jawadekar,
Department of Electrical Engineering, S. S. G. M. College of Engineering Shegaon, (M.S.), India.

G. M. Dhole,
Department of Electrical Engineering, S. S. G. M. College of Engineering Shegaon, (M.S.), India.

S. R. Paraskar.
Department of Electrical Engineering, S. S. G. M. College of Engineering Shegaon, (M.S.), India.

S. S. Jadhao,
Department of Electrical Engineering, S. S. G. M. College of Engineering Shegaon, (M.S.), India.

Please see the link here: https://stm.bookpi.org/NPER-V7/article/view/5909