Master’s in Time Series, Signals and Anomaly Detection Engineering
Online master's programme of 1,900 hours in time series, signals and anomaly detection: 12 modules covering the nature of anomalies, signal preprocessing, classical statistical methods, change detection, distance and density methods, deep learning, spectral analysis, evaluation and thresholding, diagnosis and real-time deployment, with an assigned technical tutor.
Take the controls.
Run a real engineering trade-off from this field. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Characterise the different types of anomalies in time series.
Preprocess sensor signals for analysis.
Apply statistical and change detection methods.
Develop detectors based on deep learning.
Evaluate detectors with appropriate event-based metrics.
Explain and prioritise anomalies for diagnosis.
Deploy real-time detection systems.
Who it's for
- Time-series data scientists building detectors for sensor and transaction data.
- Signal processing engineers working with vibration, audio and physiological signals.
- Process monitoring analysts tracking the health of industrial plants.
- Reliability engineers detecting early failure signatures in equipment.
- AI-driven cybersecurity analysts hunting for intrusions and abnormal behaviour.
- Control alarm engineers rationalising and prioritising plant alarms.
- Engineering or science graduates aiming to move into the process industries.
What you'll work with
- ANSI/ISA-18.2 (management of alarm systems for the process industries)
- IEC 62682 (management of alarm systems)
- ISO 7870-2 (Shewhart control charts)
- ISO 13379-1 (condition monitoring and diagnostics of machines: data interpretation)
- ISO 17359 (condition monitoring and diagnostics of machines: general guidelines)
- ISO 16079-2 (condition monitoring of wind turbines)
- Regulation (EU) 2016/679 (GDPR) (EU General Data Protection Regulation)
- Python with PyOD
- ruptures
- STUMPY
- MATLAB Signal Processing Toolbox
- Merlion
12 modules · 1900 h
1.1Point, contextual and collective anomalies1.2Novelty, outliers and regime changes1.3Anomalies in univariate and multivariate series1.4Scarce labelling and unsupervised learning1.5The cost of false positives and false negatives1.6Cases in industry, finance and networks1.7Taxonomy of detection methods1.8Benchmark datasets1.9Evaluation problems in the literature1.10Case study: defining the problem in an electricity grid
2.1Missing values and temporal imputation2.2Resampling and sensor alignment2.3Removing trend and seasonality2.4Noise filtering2.5Normalisation by operating regime2.6Time-series segmentation2.7SAX symbolic representations2.8Dimensionality reduction2.9Detecting frozen and saturated sensors2.10Lab: cleaning plant sensor data
3.1Z-scores and robust methods with MAD3.2Shewhart control charts3.3CUSUM and EWMA charts3.4ARIMA model residuals3.5STL decomposition and residuals3.6Grubbs and generalised ESD tests3.7Multivariate statistical process control3.8Principal component analysis and Q statistics3.9Assumptions and limitations3.10Practical: CUSUM detection on a process line
4.1Offline change point detection4.2The PELT algorithm and binary segmentation4.3Online Bayesian methods4.4The likelihood ratio test4.5Changes in mean, variance and spectrum4.6Sensor drift detection4.7Penalties and model selection4.8Changes in multivariate series4.9Validating change points4.10Lab: change points with the ruptures library
5.1Nearest neighbours and k-NN distance5.2The local outlier factor (LOF)5.3Isolation Forest and isolation trees5.4One-Class SVM with Gaussian kernels5.5Matrix Profile and discords5.6Clustering with DBSCAN5.7DTW distances between subsequences5.8Scalability to large volumes5.9Hyperparameter selection without labels5.10Practical: Matrix Profile for ECG anomalies
6.1Autoencoders and reconstruction error6.2Variational autoencoders and reconstruction probability6.3Forecasting LSTMs and residual error6.4Generative adversarial networks6.5Transformers for anomalies6.6Graph neural networks for interrelated sensors6.7Contrastive learning for normal representations6.8Scoring and thresholding6.9Robustness to training contamination6.10Lab: LSTM autoencoder on telemetry
7.1Spectra and characteristic bands7.2The short-time Fourier transform7.3Continuous and discrete wavelets7.4Empirical mode decomposition7.5Envelope analysis7.6Cepstrum and periodicity detection7.7Coherence between channels7.8Spectral features for detectors7.9Fault signatures in machinery7.10Workshop: spectral detection of bearing faults
8.1Point-wise versus event-based metrics8.2Precision, recall and adjusted F18.3ROC and PR curves8.4Range-based metrics and time to detection8.5Threshold selection without labels8.6Extreme value theory for thresholds8.7Validation with injected anomalies8.8Fair comparison of methods8.9Interpretation for domain experts8.10Practical: evaluation protocol on a public benchmark
9.1Attributing anomalies to variables9.2Contributions in PCA and T²9.3SHAP for detectors9.4Classifying anomaly types9.5From warning to diagnosis9.6Expert rules and hybrid systems9.7Feedback loop with operators9.8Alarm management according to ISA-18.29.9Alarm rationalisation9.10Case study: alarm prioritisation in a chemical plant
10.1Online and streaming detection10.2Sliding windows and latency10.3Incremental model updating10.4Concept drift in detectors10.5Integration with SCADA and alarm systems10.6Detection at the edge10.7Detection service architecture10.8Monitoring the detector10.9Testing with replayed data10.10Workshop: streaming detection service
11.1Cybersecurity and network intrusions11.2Transaction fraud11.3Infrastructure monitoring11.4Healthcare and physiological signals11.5Quality in industrial processes11.6Satellite and aircraft telemetry11.7Water distribution networks11.8Structural health monitoring of bridges11.9Lessons from real projects11.10Case study: leak detection in a water network
12.1Defining the case and the data12.2Exploratory analysis of the signals12.3Preprocessing and features12.4Comparison of detectors12.5Thresholding and evaluation12.6Explaining the anomalies12.7Deployment proposal12.8Conclusions and deployment recommendations12.9Thesis report12.10Master's Thesis: defence before a panel
How you'll learn
100% online training on the SEIUM campus, with downloadable materials, recorded classes and an assigned technical tutor. There are no fixed timetables: you progress at your own pace within each intake, so the programme fits around your job.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is Python with PyOD, ruptures, STUMPY, MATLAB Signal Processing Toolbox and Merlion, in free, academic or evaluation versions. Real procedures are studied, such as fault detection in vibration signals, identifying transaction fraud and leak detection in distribution networks. Assessment is continuous by module, plus the final Master's Thesis.
Standards covered: ANSI/ISA-18.2, IEC 62682, ISO 7870-2, ISO 13379-1, ISO 17359, ISO 16079-2, Regulation (EU) 2016/679 (GDPR).
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Where this takes you
Requirements
- University degree in engineering, science or architecture, or equivalent demonstrable professional experience.
- Internet connection and a computer capable of running the programme software.
- English sufficient to read technical standards.
Ready to start?
Send your application today. Our admissions team replies within 72 hours with your study plan and funding options.
Frequently asked questions
The programme totals 1,900 hours over 12–14 months. You set your own pace within each intake, and your tutor helps you plan the workload around your job.
You work with Python with PyOD, ruptures, STUMPY, MATLAB Signal Processing Toolbox and Merlion, using free, academic or evaluation versions. Your tutor indicates the free alternative wherever one exists.
A SEIUM Institute diploma, awarded on passing the continuous assessment and the Master's Thesis. It is SEIUM's own qualification, not an official state degree; its recognition depends on the programme completed and the regulations of each country.
Yes. Modules passed in a course or postgraduate diploma in the same area are recognised in the higher programme. Ask about the recognition policy when you enrol.
A university degree in engineering, science or architecture (or equivalent professional experience), a computer able to run the software and English sufficient to read technical standards.
Start dates are set by SEIUM's faculty for each intake. Our admissions team will confirm the next available date when you apply.
Related programmes
Master’s in Data, Real-Time and Platforms Engineering
Online master's programme of 1,900 hours in data engineering, real-time processing and platforms: 12 modules covering modern data architectures, batch and streaming processing, orchestration, data quality…
Master’s in Multimodality Engineering
Online master's programme of 1,900 hours in multimodal AI engineering. 12 modules covering modality fusion and alignment, per-modality representations, transformer architectures, contrastive learning, multimodal language models,…
Master’s in Causality and Experimentation Engineering
Online master's programme of 1,900 hours in causality and experimentation: 12 modules covering causal frameworks, causal graphs, classical design of experiments, online A/B testing, complex experiments,…
More in Data, AI, Simulation & Analytics
All Data, AI, Simulation & Analytics programmes → · How to apply · Fees & financing · Scholarships