Intelligent Systems / Artificial Intelligence

Labs

Topic	Description	Notes	Quiz	Slides
00Introduction	Who we are and a map of the course: search, machine learning and language models	Learn →	Quiz →	Coming soon
01Uninformed Search	Basic definitions, state spaces, and blind search strategies (BFS, DFS, UCS, IDS)	Learn →	Quiz →	Present →
02Informed Search	Heuristic functions, greedy search, A* algorithm, and optimization techniques	Learn →	Quiz →	Present →
03Constraint Satisfaction	CSP fundamentals, backtracking, arc consistency (AC-3), and min-conflicts heuristics	Learn →	Quiz →	Present →
04Genetic Algorithms	Evolutionary search, selection mechanisms, crossover, mutation, and metaheuristic optimization	Learn →	Quiz →	Present →
05Games	Adversarial search, game trees, minimax, and alpha-beta pruning	Learn →	Quiz →	Coming soon
06Introduction to ML	Core concepts, bias and variance, regression and classification, KNN, SVM, Naive Bayes, decision trees, and K-Means	Learn →	Quiz →	Coming soon
6.5Linear & Logistic Regression	Cost function, gradient descent and the closed form, R² and RMSE, logit and sigmoid, cross-entropy, one-vs-rest and softmax	Learn →	Quiz →	Coming soon
07K-Nearest Neighbours	Global vs local, instance-based vs model-based and lazy vs eager learning, distance metrics, choosing K, KNN classification and regression	Learn →	Quiz →	Coming soon
08Naive Bayes Classifier	Bayes' theorem, conditional independence, spam filtering, the zero-probability problem and Laplace smoothing, Gaussian Naive Bayes	Learn →	Quiz →	Coming soon
09Decision Trees	Classification and regression trees, recursive construction, entropy as expected surprise, information gain, ID3 in Python, overfitting, pre- and post-pruning	Learn →	Quiz →	Coming soon
10K-Means and Clustering	Unsupervised learning, the K-means algorithm and its cost, local optima and random restarts, the elbow method and silhouette analysis, K-means++, soft K-means, hierarchical clustering and linkages, cluster evaluation metrics	Learn →	Quiz →	Coming soon
11Neural Networks	The artificial neuron and activation functions, logic gates and XOR, multilayer perceptrons and universality, gradient descent, perceptron learning, back-propagation, handwritten digit recognition	Learn →	Quiz →	Coming soon
12Transformers and Generative Models	Subword tokenization and BPE, embeddings and position, self-attention with queries, keys and values, causal masking, MLP and residual connections, logits, softmax and temperature, text generation	Learn →	Quiz →	Coming soon
13Post-training and Alignment	Pretraining vs SFT vs preference learning, reward models and RLHF, the KL constraint, reward hacking and Goodhart's law, DPO, human and AI feedback, outcome vs process supervision, verifiable rewards, evaluation	Learn →	Quiz →	Coming soon
14Evaluation, Metrics and Data Analysis	Task vs metric vs evaluator, precision, recall and F1, BLEU, ROUGE-L and COMET, LLM-as-a-judge and its biases, reading plots, subgroup analysis, distribution shift, data leakage and contamination, uncertainty, human evaluation	Learn →	Quiz →	Coming soon

Recommended projects: take part in SemEval-2027

SemEval is the yearly international workshop on semantic evaluation: organisers publish a task with data and a metric, teams build systems, and all systems are scored on the same hidden test set.

A ready-made problem: the data, the evaluation and the leaderboard come from the organisers, so you spend your time on the system and its analysis.
Compete with the world: your results are compared with teams from universities and companies everywhere.
Publish a paper: teams can write a system description paper; SemEval papers appear in the ACL Anthology. Example: Elitsa's SemEval-2026 Task 13 paper with Violeta Kastreva.
Course project: we recommend a SemEval-2027 task as your project, in teams of at most 2.

The 11 tasks

1RETECO: Reasoning-Oriented Retrieval 2Directional-Consistent Fine-Grained NLI (DiCo-NLI) 3Semantic Change Across Multiple Time Periods 4Par-PARSEME: Paraphrasing of Idioms 5ComPartMENT: Compound and Particle Verb Compositionality 6Culturally Grounded Spoken and Visual Question Answering 7CLaS: Cross-Lingual Alignment and Steering of LLMs 8StereoQueerEval: Stereotypes and Hate Speech towards LGBTQIA+ 9Analysis of Multimodal Framing in the Newsorganised by us 10AgentRisk: Risk in Agentic Multi-Turn Interactions 11VAKRA-Advanced: Multi-Hop Multi-Source Tool Calling

Task 9: Analysis of Multimodal Framing in the News

We are among the organisers of Task 9: Dr. Dimitar Dimitrov, Elitsa Yotkova and Prof. Ivan Koychev, together with Giovanni Da San Martino, Preslav Nakov, Jakub Piskorski and colleagues from Padova, MBZUAI, Helsinki, Porto, Beira Interior and ETH Zurich. It builds on the SemEval-2023 task on persuasion techniques, framing and news genre.

Framing is choosing which aspects of an issue to make salient, so that a reader takes away a particular meaning. Images can reinforce, complement or contradict the framing of the text.

Paragraph

Policy prescription and evaluationEconomicPolitical

Task 9 page → All SemEval-2027 tasks →