K Knowledge HF-D7.7-004 cross-sector 9 umiejętności
Machine learning fundamentals
Knows the main forms of machine learning, the role of training data and concepts such as overfitting, validation and model performance in plain language.
W jaki sposób serwis hrmforce to mierzy
- Metoda oceny
- Knowledge test · rho 0.40 (SD 0.13)
- hrmforce narzędzie
- Knowledge test (client-specific)
- Możliwość szkolenia
- high
- Prognoza popytu na lata 2026–2030
- rising
Test of declarative job knowledge, usually assembled per client.
Kotwice behawioralne
| Poziom | Zachowanie na tym poziomie |
|---|---|
| N1 Guided | States the difference between supervised and unsupervised learning and which data a model needs to learn. works under supervision and follows instruction · routine, one variable at a time · own task |
| N3 Proficient | Explains how a model is trained and validated and judges whether reported performance figures are credible. sets own approach and seeks input proactively · several variables, some ambiguity · own team or process |
| N5 Leading | Sets requirements for model development in the organisation and reviews third party models on design and evidence. sets the standard and the policy · strategic, under high uncertainty · organisation, value chain or profession |
N2 i N4 celowo nie zostały przypisane do konkretnych punktów odniesienia. Oceniający umieszczają je pomiędzy punktami odniesienia, zgodnie z konwencją O*NET.
Podstawowe umiejętności
Umiejętności te dziedziczą ścieżkę oceny oraz punkty odniesienia behawioralne tej konstrukcji.
| T | Umiejętności | Definicja | Prognoza popytu na lata 2026–2030 |
|---|---|---|---|
| K | Knowing the main forms of machine learning Supervised learning · Machine learning | Distinguishes supervised, unsupervised and reinforcement learning and names a practical example of each. | rising |
| K | Splitting training and validation sets Train test split · Cross validation | Splits data into training, validation and test parts so model performance is measured on unseen data. | rising |
| K | Recognising overfitting Overfitting | Recognises that a model memorises the training data and therefore performs poorly on new cases. | rising |
| K | Measuring model performance Precision · Recall · Model evaluation | Chooses suitable measures such as precision, recall and area under the curve and explains what they mean. | rising |
| K | Distinguishing classification and regression Classification · Regression | Determines whether a problem asks for a class or a number and chooses the matching model type accordingly. | rising |
| V | Engineering features Feature engineering | Turns raw data into usable predictive variables and prevents future information leaking into the model. | rising |
| K | Knowing neural networks at a high level Neural networks · Deep learning | Explains how layers, weights and training relate and why such models are hard to explain. | rising |
| V | Building an evaluation set Evaluation set · Benchmark | Assembles a fixed set of test cases with expected answers to assess models and prompts comparably. | rising |
| K | Understanding model explainability Explainable AI | Explains which techniques show which variables drive a prediction and what their limitations are. | rising |