This paper summarizes the current state of the art and recent trends in software engineering economics. It provides an overview of economic analysis techniques and their applicability to software engineering and management. It surveys the field of software cost estimation, including the major estimation techniques available, the state of the art in algorithmic cost models, and the outstanding research issues in software cost estimation.

Software engineering economics

Landslide susceptibility assessment in the Hoa Binh province of Vietnam: A comparison of the Levenberg–Marquardt and Bayesian regularized neural networks

A committee neural network for prediction of normalized oil content from well log data: An example from South Pars Gas Field, Persian Gulf

Process-based models have been widely used for storm surge predictions, but their high computational demand is a major drawback in some applications such as rapid forecasting. Few efforts have been made to employ previous databases of synthetic/real storms and provide more efficient surge predictions (e.g. using storm similarity of an individual storm to those in the database). Here, we develop an alternative efficient and robust artificial intelligent model, which predicts the peak storm surge using the tropical storm parameters: central pressure, radius to maximum winds, forward velocity, and storm track. The US Army Corp of Engineers, North Atlantic Comprehensive Coastal Study, has recently performed numerical simulations of 1050 synthetic tropical storms, which statistically represent tropical storms, using a coupled high resolution wave–surge modeling system for the east coast of the US, from Cape Hatteras to the Canadian border. This study has provided an unprecedented dataset which can be used to train artificial intelligence models for surge prediction in those areas. While numerical simulation of a storm surge at this scale and resolution (over 6 million elements scaling from 20 m to more than 100 km) is extremely expensive, the artificial intelligence takes the advantage of the previous simulations, and effectively learns the relationship between storm parameters representing storm forcing and surge. The artificial neural network method which was used for this study, was shown to outperform support vector machine for extreme storms. ANN model, which is based on a neurobiological analogy, can be conveniently developed, retrained by new data, and is nonparametric. The AI model, which was developed for Rhode Island, was validated using a set of randomly selected synthetic storms as well as real tropical storms in this region. The model performance was found satisfactory with root-mean-square error of <35 cm for observed and synthetic storms. It was also shown that it is not possible to develop a reliable artificial intelligence model for this region using a limited number of data (e.g. 200 storms), which is usually available in historical records.

An efficient artificial intelligence model for prediction of tropical storm surge

Bayesian regularization of artificial neural networks (BRANNs) were used to predict body mass index (BMI) in mice using single nucleotide polymorphism (SNP) markers. Data from 1896 animals with both phenotypic and genotypic (12 320 loci) information were used for the analysis. Missing genotypes were imputed based on estimated allelic frequencies, with no attempt to reconstruct haplotypes based on family information or linkage disequilibrium between markers. A feed-forward multilayer perceptron network consisting of a single output layer and one hidden layer was used. Training of the neural network was done using the Bayesian regularized backpropagation algorithm. When the number of neurons in the hidden layer was increased, the number of effective parameters, γ, increased up to a point and stabilized thereafter. A model with five neurons in the hidden layer produced a value of γ that saturated the data. In terms of predictive ability, a network with five neurons in the hidden layer attained the smallest error and highest correlation in the test data although differences among networks were negligible. Using inherent weight information of BRANN with different number of neurons in the hidden layer, it was observed that 17 SNPs had a larger impact on the network, indicating their possible relevance in prediction of BMI. It is concluded that BRANN may be at least as useful as other methods for high-dimensional genome-enabled prediction, with the advantage of its potential ability of capturing non-linear relationships, which may be useful in the study of quantitative traits under complex gene action.

Prediction of body mass index in mice using dense molecular markers and a regularized neural network.

It is a well known fact that at the beginning of any project, the software industry needs to know, how much will it cost to develop and what would be the time required ? . This paper examines the potential of using a neural network model for estimating the lines of code, once the functional requirements are known. Using the International Software Benchmarking Standards Group (ISBSG) Repository Data (release 9) for the experiment, this paper examines the performance of back propagation feed forward neural network to estimate the Source Lines of Code. Multiple training algorithms are used in the experiments. Results demonstrate that the neural network models trained using Bayesian Regularization provide the best results and are suitable for this purpose.

Bayesian Regularization in a Neural Network Model to Estimate Lines of Code Using Function Points

Software maintenance is a task that every development group has to face when the software is delivered to the customers' site, installed and is operational. The time spent and effort required for keeping software operational consumes about 40-70% of cost of entire life cycle. This study proposes a four parameter integrated measure of software maintainability using a fuzzy model. The study also includes empirical data of maintenance time of projects which has been used to validate the proposed model.

Measurement of Software Maintainability Using a Fuzzy Model

It is well known that soft computing techniques can be very well deployed for software engineering applications. Among these fuzzy and neural models are widely used to estimate lines of codes, effort, software maintainability, software understandability etc. This paper proposes to carry out a sensitivity analysis of the two models and shows which one is better. This is done with the help of a case study where the two models are used to measure software maintainability.

Sensitivity analysis of fuzzy and neural network models

Software Engineering as a discipline emerged in response to the software crisis perceived by the industry. It is a well known fact that at the beginning of any project, the software industry needs to know how much will it cost to develop and what would be the time required. Resource estimation in software engineering is more challenging than resource estimation in any other industry. A number of resource estimation methods are currently available and the neural network model is one of them. This paper proposes to evaluate various training algorithms in a neural network model and shows which is the best suited for software engineering applications.

Evaluation of various training algorithms in a neural network model for software engineering applications

Resource Estimation is a challenging activity, in the early stages of project development. Once the functionality desired by the user is ascertained, function points can be calculated. This paper proposes to estimate Lines of Code once the Function Point count is known, using linear regression techniques and also a neural network model. These two are then combined to propose an expert committee model which gives better results. This is validated by empirical data available from ISBSG data set (release 9).

https://dl.acm.org/doi/pdf/10.1145/1095430.1095439

Manimala Puri

Papers

Bayesian Regularization in a Neural Network Model to Estimate Lines of Code Using Function Points

Measurement of Software Maintainability Using a Fuzzy Model

Sensitivity analysis of fuzzy and neural network models

Evaluation of various training algorithms in a neural network model for software engineering applications

An expert committee model to estimate lines of code