Bindu Garg

A new computational fuzzy time series model to forecast number of outpatient visits

Forecasting number of outpatient visits is pre-eminent for patient planning, medical resource utilization and overall management of health care system to a certain extent. Aim of forecasting the outpatient visits can also be seemed in terms of individual care. In addition, accurate prediction of outpatient visits in hospitals can play a significant role in health insurance plans and for deciding reimbursement system. As such, main challenge in healthcare simulation is to produce a realistic model that must utilize efficient techniques for managing complex time series data and should be capable of generating forecasted value with almost negligible error. We proposed forecasting model based on fuzzy time series that rectifies the existing imperfections and overcome the drawbacks of previous approaches. Novice concept introduced to eliminate the inadequacies by way of defining the universe of discourse on historical data. Model also endeavors to pontificate the issue of improving forecasting accuracy through the new idea of event discretization function. This was quite encouraging as it highlights the impact of trend & seasonal components by yielding dynamic change of values from time t to t+1. This fuzzy computing time series model is designed by joint consideration of three key points (1) Event discretization of time series data (2) Frequency density based partitioning (3) Creation of Fuzzy logical relationships in optimized way. Subsequently, performance of the proposed model is demonstrated and compared with some of the pre-existing forecasting methods on same outpatient data. In general, findings of the study are interesting and superior in terms of least Average Forecasting Error Rate (AFER) and Mean Square Error (MSE) values.

Fuzzy time series model to forecast rice production

Crop production is considered as one of the real world complex problem due to its non-deterministic nature and uncertain behavior. Particularly, forecasting of rice production for a lead year is pre-eminent for crop planning, agro based resource utilization and overall management of rice production. As such, main challenge in rice production forecasting is to generate realistic method that must be capable for handling complex time series data and generating forecasting with almost negligible error. The objective of present work is to design & implement such a competent fuzzy time series model for forecasting of rice production. We have proposed forecasting model based on fuzzy time series that highlights the impact of trend & seasonal components by yielding dynamic change of values from time t to t+1. The aim of using fuzzy time series is to deal with forecasting under the fuzzy environment that contains uncertainty, vagueness and imprecision. This method assigns importance to fuzzy intervals on the basis of frequency of number of time series data. Subsequently, computed fuzzy logical relations are used for analysis of time series rather than random and non-random functions as in case of usual time series analysis. Performance of the proposed model is demonstrated and compared with few pre-existing forecasting methods on rice production. To prove robustness and accuracy of the presented model, analysis is performed on forecasting of enrollment data of university of Alabama.

Enhanced accuracy of fuzzy time series predictor using genetic algorithm

Accuracy is one of the most important aspects in the domain of forecasting. It is very difficult to improve accuracy of prediction system where prediction is based only on large historical values and accuracy is important for each predicted value along with the whole system. The main objective of this research is to optimize dominant factors of fuzzy time series predictor (FTSP) using genetic algorithm (GA) and further to improve prediction accuracy for each time series variable along with whole system. This is obtained by (a) generating wide range of parameters for membership function at time t on the basis of their base value (b) subset of population generated at time t is used for fitness checking. Additionally, GA complexity is also reduced by utilizing rate of change of time series data to cut down the bit size of chromosome. It can be observed from comparative study that use of GA considerably reduced mean square error (MSE) and average forecasting error rate (AFER).

Soft Computing Model to Predict Average Length of Stay of Patient

Forecasting the average Length of Stay (LoS) of a patient is prime aspect for all hospitals to effectively determine and plan services demanded at various level. Prediction of LoS plays a vital role in strategic decision making by health care administrators. In this paper, a dynamic computational model based on time series, implemented using soft computing techniques is presented to forecast average length of stay of patient. Aim of designing proposed model is to overcome the drawbacks of the exiting approaches and derive more robust and accurate methodology to forecast LoS of patient. Subsequently, the performance of the proposed model is demonstrated by comparing the results of proposed model with some of the pre-existing forecasting methods. In general, the findings of the study are interesting and superior in terms of least Average Forecasting Error Rate (AFER) and Mean Square Error (MSE) values.

Enhanced accuracy of fuzzy time series model using ordered weighted aggregation

Display Omitted Ordered weighted aggregation (OWA) based Fuzzy time series model is proposed.Priority matrix is designed by employing regularly increasing monotonic (RIM) quantifiers.Impact of order of model and OWA weights is studied.The performances comparison in terms of least value of MSE, AFER has been realized.Robustness of proposed method has been checked. Accuracy is one of the most vital factors when dealing with forecast using time series models. Accuracy depends on relative weight of past observations used to predict forecasted value. Method of aggregation of past observations is significant aspect in time series analysis where determination of next observation depends only on past observations. Previous research on fuzzy time series for forecasting treated fuzzy relationship equally important which might not have properly reflected the importance of each individual fuzzy relationship in forecasting that introduced inaccuracy in results. In this paper, we propose ordered weighted aggregation (OWA) for fuzzy time series and further design forecasting model signifying efficacy of the proposed concept. Objective of using fuzzy time series is to deal with forecasting under the fuzzy environment that contains uncertainty, vagueness and imprecision. OWA is utilized to generate weights of past fuzzy observations; thereby eliminating the need for large number of historical observations required to forecast value. OWA weights are determined by employing regularly increasing monotonic (RIM) quantifiers on the basis of fuzzy set importance using priority matrix. Experimental study reveals how OWA coalesced with fuzzy time series for designing of forecasting model. It can be observed from comparative study that use of OWA considerably reduces mean square error (MSE) and average forecasting error rate (AFER). Robustness of proposed model is ascertained by demonstrating its sturdy nature and correctness.

Employing genetic algorithm to optimize OWA-fuzzy forecasting model

Accuracy of forecasting in fuzzy based prediction system considerably depends on subjectively decided parameters such as fuzzy membership function. In this paper, we presented a novice concept to optimize Ordered Weight Aggregation (OWA) based forecasting model by Genetic Algorithm. Firstly, OWA weights are determined on the basis of importance of fuzzy set in the system by employing regularly increasing monotonic (RIM) quantifiers. Subsequently, genetic algorithm is employed to generate wide range of parameters for fuzzy membership functions (mf) in the region of time series. Lastly, forecasted value is obtained by OWA aggregation of past fuzzy observations generated at prior time (t, t−1, t−2). Proposed optimized forecasting model has been compared with some pre-existing models on same data. Results demonstrate that forecasting performance of the proposed model has greatly improved by reducing mean square error (MSE) and mean absolute percentage error (MAPE).

Fuzzy Time Series Prediction Model

The main objective to design this proposed model is to overcome the drawbacks of the exiting approaches and derive more robust & accurate methodology to forecast data. This innovative soft computing time series model is designed by joint consideration of three key points (1) Event discretization of time series data (2 Frequency density based partitioning (3) Optimizing fuzzy relationship in inventive way. As with most of cited papers, historical enrollment of university of Alabama is used in this study to illustrate the new forecasting process. Subsequently, the performance of the proposed model is demonstrated by making comparison with some of the pre-existing forecasting methods. In general, the findings of the study are interesting and superior in terms of least Average Forecasting Error Rate (AFER) and Mean Square Error (MSE) values.

Recognition of Table Images Using K Nearest Neighbors and Convolutional Neural Networks

The objective of this research paper is to analyze images of tables and build a prediction system capable of recognizing the number of rows and columns of the table image with the help of Convolutional Neural Networks and K Nearest Neighbours. The data set used in the building of the models has been indigenously created and converted to gray-scale. The eventual objective and possible application of the paper is to assist the building of software capable of reading tables from non digital sources and creating digital copies of them.

Application of Classification Techniques for Prediction of Water Quality of 17 Selected Indian Rivers

Objective: In this study, prediction using classification techniques are used to predict the water quality of the 17 selected rivers in the year 2011 using their water quality in 2008 to interpret whether the water quality has improved or deteriorated. Methods/Analysis: For this prediction, we have used data mining classification techniques using Waikato Environment for Knowledge Analysis (WEKA) API to the dataset of selected 17 Indian rivers. The data used for prediction was created from ambient water quality of Aquatic Resources in India in 2008 and 2011. Data is obtained from data portal which was published under National Data Sharing and Accessibility Policy (NDSAP) and the contributor was Ministry of Environment and Forests Central Pollution Control Board (CPCB). Findings: Out of the four techniques used, prediction of classes, i.e. excellent, good, average and fair is best done by Naive Bayes followed by J48, SMO and REPTree technique.

Non-invasive Anaemia Detection by Analysis of Conjunctival Pallor

Anaemia is a condition wherein the number of Red Blood Cells (RBCs) in the body diminishes. This leads to a decline in the oxygen-carrying capacity of the blood. Though iron deficiency is one of the prominent factors that cause anaemia, other factors like malaria, improper diet, infections, etc. are also responsible for causing anaemia. The diagnosis of anaemia often commences with a clinical inspection. This is followed by laboratory tests like Complete Blood Count (CBC), reticulocyte count, etc. These tests fall under the invasive category as they necessitate injecting surgical instruments into the patient’s body. The non-invasive methods rely on checking for pallor in regions like tongue, conjunctiva, nailbed, etc. The limitation of this technique is that it is not too reliable in cases of mild anaemia. This dissertation aims to develop a reliable, non-invasive way to diagnose anaemia. A smartphone camera is used to capture the conjunctival region. Thereafter, the image is processed and an Erythema Index (EI) is generated. The EI is used to discern if a person is anaemic or not. The experimental results indicate that the devised technique is able to identify mildly anaemic individuals to a decent level of accuracy.