Muhammad Irfan Memon

Dr. Muhammad Irfan Memon


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 January 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
SASyed Rizwan Ali
Syed Rizwan Ali
SHShaik Javeed Hussain
Shaik Javeed Hussain
MMMuhammad Irfan Memon
Muhammad Irfan Memon
MMMohammad Mohatram
Mohammad Mohatram
MFMuhammad Faraz
Muhammad Faraz
Entrepreneurship education is a key engine to equip students with the high-level skills of creativity, problem-solving and strategic decision-making to navigate the constantly evolving global economy. The current study explores the role of digital tools in enhancing entrepreneurship training, focusing on the communicative platform, the collaborative platform, the simulation platform and the gamification platform. Tools such as Moodle, Capsim, Tableau and Kahoot make experiential learning possible, connecting the abstract with the concrete. Although digital tools have a mediated impact on learning outcomes related to entrepreneurship by promoting skill development, problems with digital skills, a lack of infrastructure and a culture of resistance to change are obstacles to achieving the full playing capacity of digital tools. The liberating possibility for the potential of future technology, especially AI and VR, for providing highly personalized and immersive education is demonstrated to be meaningful to the need for equitable access to resources and for having robust change management strategies. This research highlights the key and important role of digital technologies in entrepreneurial education reform and identifies the roadblocks to integrating those technologies effectively as well as suggests intervention strategies for better quality and access to education services in all SES levels.

28 October 2026 in An International Journal of Optimization and Control: Theories & Applications (IJOCTA)
Verified
Journal Article
Open access
PDF file
SASyed RIzwan Ali
Syed RIzwan Ali
MMMuhammad Irfan Memon
Muhammad Irfan Memon
MMMohammad Mohatram
Mohammad Mohatram
MFMuhammad Faraz
Muhammad Faraz
MKMuhammad Mumtaz Khan
Muhammad Mumtaz Khan
JHJaveed Hussain
Javeed Hussain
Since Industry 4.0 entered international policy and scholarly debate in 2011, the convergence of cyber-physical systems, the Internet of Things, cloud computing, data analytics, artificial intelligence, and intelligent automation has reshaped how university-linked startups are conceived, incubated, and scaled. This study evaluates an integrated partial least squares structural equation modeling (PLS-SEM), artificial neural network (ANN), PLSpredict/cross-validated predictive ability test (CVPAT), and importance–performance map analysis (IPMA) framework linking technological infrastructure (TI), innovation capacity (IC), entrepreneurship education (EE), digital capability (DC), and Industry 4.0 readiness (I4R). A quantitative cross-sectional survey using criterion-based purposive sampling produced 250 usable responses from student entrepreneurs, incubator resident venture representatives, faculty startup advisers, and innovation center employees. PLS-SEM with 5,000 bootstrap resamples showed that DC had the strongest positive association with I4R (β = 0.485, p < 0.001), followed by TI (β = 0.241, p < 0.001). EE showed a small, marginally significant positive association (β = 0.095, p = 0.047), whereas IC was not significant (β = 0.083, p = 0.093). The model explained 53.8% of the variance in I4R (R2 = 0.538). PLSpredict yielded positive Q2predict values for all I4R indicators, and construct-level CVPAT showed significantly lower PLS-SEM prediction loss than the linear-model benchmark (difference = −0.059, t = 2.602, p = 0.010). A fully specified, repeated cross-validated multilayer perceptron ANN ranked DC first (100% normalized importance), followed by TI (47.6%), EE (10.9%), and IC (10.0%); however, the matched linear benchmark achieved lower held-out root mean square error, so ANN is interpreted as a complementary ranking tool rather than evidence of superior predictive accuracy. IPMA identified DC (importance = 0.485; performance = 48.222) and TI (0.241; 46.037) as the leading intervention priorities. The framework therefore integrates explanatory assessment, out-of-sample prediction, predictive ranking, and intervention prioritization without implying causal proof from cross-sectional data.

30 April 2026 in Journal of Visualized Experiments
Verified
Journal Article
Closed access
HSHina Shakir
Hina Shakir
MMMohammad Mohatram
Mohammad Mohatram
SHShaik Javeed Hussain
Shaik Javeed Hussain
SASyed Rizwan Ali
Syed Rizwan Ali
MMMuhammad Irfan Memon
Muhammad Irfan Memon
Radiomics enables the extraction of quantitative imaging biomarkers from medical images and has become an important tool for computer-aided cancer diagnosis. However, radiomics datasets are typically high-dimensional with limited sample sizes, making feature selection a critical step for building reliable predictive models. This study proposes a gradient-loss recursive feature elimination (GL-RFE) framework that integrates gradient sensitivity analysis from a deep neural network to identify the most influential radiomic features for lung cancer stage detection. A total of 106 radiomic features were extracted from chest computed tomography (CT) scans using the PyRadiomics extension of the 3D Slicer platform. The proposed method evaluates feature importance by computing gradients of the network loss with respect to input features and recursively eliminates features with minimal contribution. The resulting top 15 radiomic features are used to train a deep neural network classifier for distinguishing early-stage and advanced-stage lung cancer. The proposed framework achieves strong classification performance, with an accuracy of 90.22%, precision of 90.10%, recall of 90.24%, and F1-score of 90.16% on the test dataset. Visualization analyses, including correlation heat maps and distribution plots, further confirm reduced feature redundancy and improved class separability. Compared to conventional feature selection techniques, GL-RFE effectively captures nonlinear feature interactions and enhances model generalization. The presented protocol provides a reproducible and interpretable methodology for radiomics-based cancer stage detection. It is particularly suitable for high-dimensional, small-sample biomedical datasets and has potential applications in other domains, such as genomics and multimodal clinical analysis.