EXIN BCS ML Award

30 minutes for 18 multiple-choice questions. Two of those are scenario-based questions worth 2 points each, so the scenario items carry double weight. That works out to roughly 100 seconds per question on average, which is tight enough that you should not plan to deliberate long on any single item.

The published pass mark is 65%. Because the two scenario-based questions are worth 2 points each, they carry disproportionate weight relative to the single-point items, so accuracy on the scenario items matters more than raw question count suggests. Aim comfortably above the threshold in practice runs rather than just at it.

No. EXIN lists the exam as not open book and states that electronic equipment is not allowed. Plan to work from memory: the algorithms, the learning types, the tool names and the data-preparation steps need to be recallable without notes. Practise answering without referring to the reading list.

EXIN lists English and Portuguese as the available languages. If you are working from the English reading list but sitting the Portuguese paper, check that the terminology you have learned maps onto the translated terms for concepts such as supervised learning, clustering and validation before exam day.

It is a foundation-level award, and EXIN explicitly markets it as suitable for people new to AI and for those with little programming experience. You need conceptual fluency rather than mathematical derivation: what each algorithm does, when it is used, how data is prepared, and how models are trained and evaluated.

Four published subject areas: Introduction to Machine Learning (definitions, applications, learning agents, deep learning, neural networks, knowledge-based system integration, data interaction); Programming in Machine Learning (languages, open source versus proprietary tools); Machine Learning Algorithms (mathematical foundations, common algorithms, supervised, unsupervised and semi-supervised learning); and Practical Applications (problem identification, data preparation, training, testing and validation, reporting to stakeholders).

EXIN names required reading: Géron's Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow; Theobald's Machine Learning for Absolute Beginners; Strang's Linear Algebra and Learning from Data; Lowe and Lawless's Artificial Intelligence and Machine Learning Foundations; and Burnett's AI in Business. If you are new, start with Theobald for vocabulary, then Lowe and Lawless for the exam's conceptual frame, using Géron for tools and Strang only for the mathematical foundations section.

Yes. The certification page offers a Preparation Guide, a downloadable sample exam and an online sample exam. Use the sample to calibrate your pace against the 30-minute limit and to see how the scenario-based format is phrased, since those two items are worth double and reward reading the whole scenario before choosing.

Three recurrences: confusing supervised, unsupervised and semi-supervised learning when a scenario describes labelled versus unlabelled data; mixing up classification with regression or clustering; and treating data preparation as a minor step when the exam treats collection, cleaning, preprocessing and transformation as a subject area in its own right. Also expect questions on bias and ethical and legal concerns, which candidates often skip.

The source page does not state a retake policy or a certification validity period, so confirm both with EXIN or your accredited training partner before booking. Do not assume a fixed resit window or a renewal cycle based on other EXIN awards; the published details for this certification list only duration, question count, pass mark, open-book status, equipment rules, level and languages.