Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jean-Claude Malela-Majika has a total of 90 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    SAGE-CUSUM: Using attention and similarity mechanisms in the construction of a new adaptive ratio control chart with a Bearing Degradation Case Study

    by Parviz Ghafariasl, Ali Yeganeh, Shing I. Chang, Jean-Claude Malela-Majika

    Monitoring the ratio of two normal variables is challenging because ratio statistics can exhibit non-normality, nonlinear sensitivity to denominator variability, and unstable behavior under noise and dependence. This study proposes SAGE-CUSUM, an adaptive ratio control chart that combines probability calibration, contextual similarity modeling, robust standardization, attention-guided enhancement, dual-timescale EWMA memory, adaptive gated fusion, and two-sided FIR-CUSUM decision logic. The framework is designed to detect sustained ratio changes while limiting unnecessary responses to temporary disturbances. SAGE-CUSUM was compared with Shewhart, distribution-free normalized CUSUM (DFNC), and TCN-based predictive monitoring under comparable in-control calibration. A Monte Carlo study covering 540 operating scenarios showed that SAGE-CUSUM achieved the strongest aggregate performance, with a mean detection rate of 0.995 and substantial detection-speed advantages for small shifts under realistic variability. This advantage was not universal: DFNC remained faster in a near-noiseless boundary regime. A paired transient-disturbance experiment further showed that SAGE-CUSUM produced the smallest mean increase in alarm probability attributable to temporary shifts (0.0247 versus 0.0766–0.0988 for the competing methods). The methods were also evaluated on 14 PRONOSTIA/FEMTO-ST run-to-failure bearing trajectories. SAGE-CUSUM and DFNC established persistent warnings at similar median remaining-trajectory fractions (0.91 and 0.90), but SAGE-CUSUM generated one coherent alarm episode per bearing with no subsequent recovery. Overall, the results indicate that SAGE-CUSUM is most valuable in noisy monitoring environments where sustained-shift sensitivity, resistance to temporary disturbances, and warning stability are simultaneous operational priorities.

  • Open Access

    A Rank‐Based CUSUM Monitoring Scheme for Time Between Events and Amplitude Data

    by Muhammad Arslan, Ali Yeganeh, Fatemeh Sogandi, Jean‐Claude Malela‐Majika, Fernanda Otilia Figueiredo, Aisha Iqbal

    Monitoring time between events and amplitude (TBEA) data is critical in many applications, particularly in manufacturing, reliability engineering, and service operations. Yet, most existing TBEA control charts focus on a single quality characteristic and are typically developed under restrictive distributional assumptions. In practice, each event is characterized by two independent variables: the waiting time since the previous event and the magnitude (amplitude) of the event itself, whose underlying distributions are often unknown. This paper proposes a novel rank‐based one‐sided cumulative sum (CUSUM) control chart for univariate TBEA monitoring, designed to jointly monitor shifts in both the interarrival time and the amplitude. The run‐length properties of the proposed nonparametric CUSUM‐TBEA chart are derived via a Markov chain approximation, enabling computation of in‐control (IC) and out‐of‐control (OOC) performance measures using closed‐form expressions. Monte Carlo simulations confirm the accuracy of the Markov chain results and provide guidance on the required simulation size for reliable performance estimation. Under identical IC conditions, the proposed CUSUM–TBEA chart generally exhibits faster detection and reduced run‐length variability compared to the classical exponentially weighted moving average (EWMA) chart, particularly for moderate to large shifts. The effects of Phase I sample size, unknown shift magnitudes, and correlation are also investigated. The results reveal robust performance across a wide range of monitoring scenarios. An illustrative example involving wildfire occurrence data further highlights the practical applicability of the proposed method. Overall, the results show that the proposed chart is a powerful distribution‐free and a practical tool for monitoring event‐driven processes with concurrent changes in event timing and magnitude.

  • Generating a combined analysis using linear mixed models to account for pre-history effects when a treatment factor comprises fixed and random levels

    by Lyson Chaka, Peter Njuho, Hans-Peter Piepho, Jean-Claude Malela-Majika

    Experimental designs involving factors with a mix of fixed and random levels have been explored by few. These designs are useful when comparing a set of new treatments (fixed levels) to a population of established treatments (random levels). This approach enables partitioning variability and testing of both fixed and random effects, leading to improved estimates and more reliable inference. However, combining analyses of variance from partitioned data poses challenges, including data rearrangement, methodology, and coding complexities. Existing statistical software does not directly support combined analyses for such designs. The current study provides guidelines for conducting combined analysis of variance in linear mixed-effects models where factors have both fixed and random levels. The approach utilizes SAS PROC GLIMMIX and tools from the Comprehensive R Archive Network to compile the combined analysis of variance, followed by multiple comparisons of treatment means. The procedure is demonstrated using treatment structures in a completely randomized design, split-split-plot design, and repeated-measures design. The method can be extended to other experimental designs. This framework addresses a critical gap in existing literature by providing a practical and readily applicable analytical tool that enables integrated analysis across all factor levels within a single linear mixed-model structure, while accounting for pre-history effects associated with prior management practices in complex experimental systems.

  • Open Access

    Phase I Multivariate Coefficient of Variation Control Charts for High‐Dimensional Processes

    by O. A. Oyegoke, K. S. Adekeye, J.‐C. Malela‐Majika

    Multivariate control charts have traditionally focused on monitoring the process mean vector and/or covariance matrix. Recent studies have extended this framework to Phase II monitoring of the multivariate coefficient of variation (MCV); yet very limited work has examined MCV‐based monitoring in Phase I. This gap is critical because Phase I procedures establish the IC state by estimating control limits from historical data, which may contain contamination or atypical observations. Robust Phase I methods are therefore essential for identifying and removing such contaminated samples before deploying Phase II monitoring. However, no existing research has evaluated the performance of MCV charts for high‐dimensional Phase I processes, where covariance estimation becomes unstable and classical methods fail. This study investigates the Phase I performance of several MCV control charts using a reduction approach built on the Least Absolute Shrinkage and Selection Operator (LASSO)‐penalized likelihood ratio framework. The proposed charts are evaluated under the localized coefficient of variation disturbance scenarios using the probability to signal () as the primary performance metric. A real‐world case study using diagnosed breast cancer data demonstrates the practical application of the proposed Phase I methodology. The results provide valuable guidance for practitioners in industrial and healthcare environments by identifying efficient and computationally stable Phase I control charts for monitoring the MCV of a process in high‐dimensional environments.


  • Open Access

    Windowed Mean Drift Exponentially Weighted Moving Average Control Chart for Monitoring Complex Autocorrelated Processes

    by Jeanette Maria Louw, Jean‐Claude Malela‐Majika, Kayode Samuel Adekeye, Aamir Saghir

    In modern manufacturing environments, traditional statistical process control (SPC) methods often struggle with complex, dynamic data patterns, particularly when observations are autocorrelated. Control charts are useful tools used in SPC to detect any significant drift in a process. Thus, there is an increasing interest in improving their detection ability, regardless of the nature and complexity of the data. This paper introduces two new exponentially weighted moving average (EWMA) charts for monitoring mean drifts in a process. The first one is named, mean drift EWMA (MD‐EWMA) chart, and the second one, named windowed mean drift EWMA (WMD‐EWMA) chart, which is the enhanced version of the MD‐EWMA chart. The effectiveness of both charts in detecting moderate and large mean drifts while minimising false positives is explored under different scenarios. Furthermore, the performance of the WMD‐EWMA chart is compared with the MD‐EWMA, EWMA and CUSUM charts. The comparison results reveal the necessity of a window‐based approach in reducing false positives, particularly when specific behavioural patterns are expected. Through extensive simulations and real‐world case studies, the WMD‐EWMA chart is shown to outperform the MD‐EWMA, Shewhart, EWMA, and CUSUM charts by effectively identifying significant mean drifts and reducing false alarms.


  • Open Access

    Optimal design of risk-based average charts for autocorrelated measurements

    by Aamir Saghir, Zahid Khan, Jean-Claude Malela-Majika, Zsolt T. Kosztyán


  • Open Access

    Process capability indices for Marshall–Olkin inverse log-logistic distribution

    by Olubisi Lawrence Aako, Kayode Samuel Adekeye, Johnson Ademola Adewara, Jean-Claude Malela-Majika

    Process capability analysis is a vital tool in quality management that enables organizations to evaluate and enhance their processes. Real-world data are mostly non-normal, they often deviate from the assumption of normality. The estimators of process capability indices (PCIs) for normal processes are not sufficient to characterize non-normal processes and can give misleading results. The Marshall-Olkin inverse log-logistic (MO-ILL) distribution is a flexible distribution that can effectively model data exhibiting positive skewness, asymmetry and heavy tails. In this paper, we derived the process capability indices (PCIs) based on the MO-ILL distribution when the process is assumed to be in a state of statistical control. Two PCIs based on MO-ILL mean and variance, and MO-ILL quantiles are proposed. The proposed PCIs were compared with the traditional PCIs and percentile-based PCIs using two real life data and data generated from MO-ILL distribution. Moreover, the effect of the sample size and parameters of the MO-ILL distribution on the PCI measures is also investigated. The results showed that PCIs values based on the proposed MO-ILL mean and variance, and MO-ILL quantiles are respectively lower and better than the traditional PCIs and percentile-based PCIs. This is an indication that MO-ILL distribution-based methods developed have narrow margin of error and are more appropriate in assessing the performance of a skewed process.