Muhammad Umer

Muhammad Ahmed Mohsin, Muhammad Umer, Ahsan Bilal et al.

Large Language Models (LLMs) have benefited enormously from scaling, yet these gains are bounded by five fundamental limitations: (1) hallucination, (2) context compression, (3) reasoning degradation, (4) retrieval fragility, and (5) multimodal misalignment. While existing surveys describe these phenomena empirically, they lack a rigorous theoretical synthesis connecting them to the foundational limits of computation, information, and learning. This work closes that gap by presenting a unified, proof-informed framework that formalizes the innate theoretical ceilings of LLM scaling. First, computability and uncomputability imply an irreducible residue of error: for any computably enumerable model family, diagonalization guarantees inputs on which some model must fail, and undecidable queries (e.g., halting-style tasks) induce infinite failure sets for all computable predictors. Second, information-theoretic and statistical constraints bound attainable accuracy even on decidable tasks, finite description length enforces compression error, and long-tail factual knowledge requires prohibitive sample complexity. Third, geometric and computational effects compress long contexts far below their nominal size due to positional under-training, encoding attenuation, and softmax crowding. We further show how likelihood-based training favors pattern completion over inference, how retrieval under token limits suffers from semantic drift and coupling noise, and how multimodal scaling inherits shallow cross-modal alignment. Across sections, we pair theorems and empirical evidence to outline where scaling helps, where it saturates, and where it cannot progress, providing both theoretical foundations and practical mitigation paths like bounded-oracle retrieval, positional curricula, and sparse or hierarchical attention.

Muhammad Umer, Robi Polikar

One of the more challenging real-world problems in computational intelligence is to learn from non-stationary streaming data, also known as concept drift. Perhaps even a more challenging version of this scenario is when -- following a small set of initial labeled data -- the data stream consists of unlabeled data only. Such a scenario is typically referred to as learning in initially labeled nonstationary environment, or simply as extreme verification latency (EVL). Because of the very challenging nature of the problem, very few algorithms have been proposed in the literature up to date. This work is a very first effort to provide a review of some of the existing algorithms (important/prominent) in this field to the research community. More specifically, this paper is a comprehensive survey and comparative analysis of some of the EVL algorithms to point out the weaknesses and strengths of different approaches from three different perspectives: classification accuracy, computational complexity and parameter sensitivity using several synthetic and real world datasets.

Muhammad Umer, Saima Sadiq, Abid Ishaq et al.

Rapid increase of traffic volume on urban roads over time has changed the traffic scenario globally. It has also increased the ratio of road accidents that can be severe and fatal in the worst case. To improve traffic safety and its management on urban roads, there is a need for prediction of severity level of accidents. Various machine learning models are being used for accident prediction. In this study, tree based ensemble models (Random Forest, AdaBoost, Extra Tree, and Gradient Boosting) and ensemble of two statistical models (Logistic Regression Stochastic Gradient Descent) as voting classifiers are compared for prediction of road accident severity. Significant features that are strongly correlated with the accident severity are identified by Random Forest. Analysis proved Random Forest as the best performing model with highest classification results with 0.974 accuracy, 0.954 precision, 0.930 recall and 0.942 F-score using 20 most significant features as compared to other techniques classification of road accidents severity.

Muhammad Junaid Umer, Qaiser Riaz

Robotics system for rehabilitation of movement disorders and motion assistance are gaining increased intention. In this scenario estimation of ground contact is an active area of research in robotics and healthcare. This article addresses the estimation and classification of right and left foot during the healthy human gait based on the IMU sensor data of chest and lower back. For this purpose we have collected an IMU data of 48 subjects by using two smartphones at chest and lower back of the human body and one smart watch at right ankle of the body. To show the robustness of our approach data was collected at six different surfaces (road tiles carpet grass concrete and soil). The recorded data of lower back and chest sensor was segmented into single steps on the basis of right ankle sensor data, then we computed a total of 408 features from time frequency and wavelet domain of each segmented step. For classification task we have trained two machine learning classifiers SVM and RF with 10 fold cross validation method. We performed classification experiments at individual surfaces, hard surfaces, soft surfaces and all surfaces, highest accuracy was achieved at individual surfaces with accuracy index of 98.88%. Furthermore, classification rate at hard soft and all surface are 95.60%, 94.38% and 95.05% respectively. The results shows that estimation of ground contact form normal human walk at different surfaces can be performed with high accuracy using 6D data of angular velocities and accelerations from chest and lower back location of the body.