RNN-CNN MODEL:A Bi-directional Long Short-Term Memory Deep Learning Network For Story Point Estimation

Marapelli, Bhaskar; Carie, Anil; Islam, Sardar M. N.

doi:10.1109/citisia50690.2020.9371770

Cited by 13 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This renders the Bi-LSTM model very effective and efficient at handling sequential dependencies between words and phrases in both directions of the sequence [23]. Marapelli et al [18] proposed an RNN-CNN model, a Bidirectional LSTM Deep Learning Network for estimating story point effort for user stories based on the historical dataset of user story descriptions along with their estimated story point effort and observed that the Bi-LSTM forward and backward feature learning allowed their model to conserve the sequence data resulting in the CNN component to produce more accurate feature extraction.…”

Section: ) Long Short-term Memory (Lstm)mentioning

confidence: 99%

“…The output from the previous layer is fed to one or more fully connected layers to produce the classification. The convolutional Neural Network, although primarily targeted towards image processing, has been effectively exploited for text classification as illustrated by [18] and [19]. Ochodek et al [17] observed that the CNN model yields better accuracy around estimating COSMIC size using use-case names as compared to the RNN and CNN-RNN approaches.…”

Section: ) Convolutional Neural Network (Cnn)mentioning

confidence: 99%

“…Based on literature, the LSTM and the RNN/RHN techniques were most used. The Bi-LSTM + CNN model proposed byMarapelli et al (2020), however, yielded very decent results. Lately, graph neural networks (GNN) have demonstrated remarkable performances around textual data classification for effort estimation as illustrated by the different studies byPhan and Jannesari (2022).…”

mentioning

confidence: 99%

See 2 more Smart Citations

A SLR on Deep Learning Models Based on Textual Information For Effort Estimation in Scrum

Sembhoo

Gobin-Rahimbux

2023

Preprint

View full text Add to dashboard Cite

Deep Learning (DL) is a branch of Machine Learning where models are developed using neural networks made of several layers for prediction. DL models have been developed to predict effort estimation in software development. This paper presents a review of works which discuss the use of DL models for effort estimation for Scrum. The various textual information, the different DL techniques used. The methodology used for the review is snowballing. It was found that Deep-SE, a model which combines LSTM and RHN has been developed specifically for effort estimation. Also, a number of other DL techniques which have been experimented are discussed. A number of performance metrics were identified and also the perfomance of the various models were compared.

show abstract

Section: ) Long Short-term Memory (Lstm)mentioning

confidence: 99%

Section: ) Convolutional Neural Network (Cnn)mentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

A SLR on Deep Learning Models Based on Textual Information For Effort Estimation in Scrum

Sembhoo

Gobin-Rahimbux

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…as well as machine learning algorithms like, "fuzzy, 43,44 DNN, 45,46 LSTM, 47 and ELM 48,49 " in terms of various error measures like, "MEP, SMAPE, rate is a hyper-parameter that controls how much we are adjusting the weights of our network concerning the loss gradient. Learning rate is a hyper-parameter that controls the weights of the network concerning the loss gradient."…”

Section: Experimental Evaluationmentioning

confidence: 99%

Automated software effort estimation for agile development system by heuristically improved hybrid learning

Gupta¹,

Mahapatra²

2022

Concurrency and Computation

View full text Add to dashboard Cite

Summary In the process of modern software development, an important role is played by software effort estimation. The failure or success of the projects is mostly based on the schedule outcomes and effort estimation accuracy. The effort estimation problem still remains as a challenging issue because of the limitations of various standard measures for forecasting the efforts in the plan‐driven software development. This article emphasizes on novel software effort estimation framework using heuristically improved hybrid learning model. This article proposes heuristically improved hybrid learning (HI‐HL) with deep belief network and artificial neural network for software effort estimation. The weight optimization strategy of DBN and ANN to attain highly accurate estimation. Here, the weight optimization is performed by integrating two standard optimization algorithms such as forest optimization algorithm and moth‐flame optimization, so called as solution index‐based forest moth flame optimization with the purpose of solving the fitness function concerning the “Magnitude of Relative Error and Mean Absolute Error (MAE).” From the table results, for dataset 1, the SMAPE of SI‐FMFO‐HI‐HL is correspondingly secured 60.75%, 34.26%, 56.77%, and 61.44% improved than ELM, LSTM, DNN, and fuzzy. The simulation findings indicated that the recommended estimation model outperformed the baseline approaches on the three publically available datasets.

show abstract

“…Gating RNN is a technique for solving the problem of disappearing gradients. Long shortterm memory (LSTM) and gated recurrent unit (GRU) are two well-known gating techniques [7], [35]- [37]. The convolutional neural network (CNN) is a sort of multi-layer neural network.…”

Section: Introductionmentioning

confidence: 99%

Multimodal video abstraction into a static document using deep learning

Abdulsahib

Abdulmunim

2023

IJECE

View full text Add to dashboard Cite

<p><span lang="EN-US">Abstraction is a strategy that gives the essential points of a document in a short period of time. The video abstraction approach proposed in this research is based on multi-modal video data, which comprises both audio and visual data. Segmenting the input video into scenes and obtaining a textual and visual summary for each scene are the major video abstraction procedures to summarize the video events into a static document. To recognize the shot and scene boundary from a video sequence, a hybrid features method was employed, which improves detection shot performance by selecting strong and flexible features. The most informative keyframes from each scene are then incorporated into the visual summary. A hybrid deep learning model was used for abstractive text summarization. The BBC archive provided the testing videos, which comprised BBC Learning English and BBC News. In addition, a news summary dataset was used to train a deep model. The performance of the proposed approaches was assessed using metrics like Rouge for textual summary, which achieved a 40.49% accuracy rate. While precision, recall, and F-score used for visual summary have achieved (94.9%) accuracy, which performed better than the other methods, according to the findings of the experiments.</span></p>

show abstract

RNN-CNN MODEL:A Bi-directional Long Short-Term Memory Deep Learning Network For Story Point Estimation

Cited by 13 publications

References 21 publications

A SLR on Deep Learning Models Based on Textual Information For Effort Estimation in Scrum

A SLR on Deep Learning Models Based on Textual Information For Effort Estimation in Scrum

Automated software effort estimation for agile development system by heuristically improved hybrid learning

Multimodal video abstraction into a static document using deep learning

Contact Info

Product

Resources

About