📖 Glossary ============ Common terms used throughout the SimBA documentation. Terms defined here can be cross-referenced from anywhere in the docs with the ``:term:`` role (e.g. ``:term:`ROI``` renders as :term:`ROI`). .. glossary:: :sorted: annotation labelling The process of marking, frame-by-frame, whether a :term:`behavior` is present or absent in a video. These human labels are the ground truth used to train a :term:`classifier`. body-part A single tracked point on an animal (e.g. ``nose``, ``left_ear``, ``tail_base``), produced by :term:`pose estimation` and stored as ``x``, ``y`` (and probability) columns. bounding box The smallest axis-aligned (or rotated) rectangle that encloses an animal or a set of body-parts; used for overlap, area and proximity computations. bout A continuous, uninterrupted episode of a behavior — i.e. a run of consecutive frames classified as the same :term:`behavior`. Bout-level statistics summarise the count, duration and timing of these episodes. behavior target The action a SimBA :term:`classifier` is trained to detect (e.g. attack, grooming, freezing). Also referred to as the *target*. circular statistics Statistics for angular/directional data (degrees), where 359° and 1° are close. Used in SimBA for heading, turning and directional analyses. classifier A supervised machine-learning model (typically a random forest) trained on annotated :term:`features ` to predict the presence of a :term:`behavior` on each frame. convex hull The smallest convex polygon enclosing a set of body-parts; a common basis for animal area, shape and overlap metrics. DeepLabCut DLC A popular open-source :term:`pose estimation` toolbox. SimBA imports DLC tracking data (single- and multi-animal). directionality Whether, and where, an animal is facing — e.g. toward another animal, a body-part or a :term:`ROI`. egocentric alignment Re-centering and rotating each frame so a chosen body-part is fixed in position and orientation, removing the animal's global location/heading from the analysis. feature feature extraction A numeric quantity computed per frame from :term:`pose estimation` data (distances, velocities, angles, areas, etc.). Feature extraction turns raw tracking into the inputs a :term:`classifier` learns from. FPS Frames per second — the video frame rate. Required to convert frame counts to seconds and to compute time-based metrics. FSTTC Forward Spike Time Tiling Coefficient — a measure of the temporal association between two behaviors (how often one tends to follow another within a time window), adapted from spike-train analysis. Gantt plot A timeline visualization showing when each :term:`behavior` occurs across a session as horizontal bars. heatmap A spatial visualization of where an animal spends time (location heatmap) or where a :term:`behavior` occurs, binned over the arena. interpolation Filling in missing body-part coordinates (e.g. dropped/occluded frames) by estimating values from neighbouring frames. Kleinberg smoothing burst detection A burst-detection algorithm (Kleinberg, 2003) applied to classifier output to merge fragmented detections into coherent :term:`bouts ` and remove noise. maDLC Multi-animal DeepLabCut — the multi-animal variant of :term:`DeepLabCut`. machine results The per-video CSV files (in ``project_folder/csv/machine_results``) holding the classifier predictions for each frame. outlier correction Detecting and correcting implausible body-part coordinates (location- and movement-based) before feature extraction. pose estimation Tracking the 2D positions of animal :term:`body-parts ` across video frames, using tools such as :term:`DeepLabCut`, :term:`SLEAP` or YOLO. project config The ``project_config.ini`` file at the root of a SimBA project, storing all project settings (paths, body-parts, classifiers, thresholds). px/mm pixels per millimeter The conversion factor between image pixels and real-world millimetres, used to report distances/speeds in physical units. Set per video via a known reference length. ROI Region of Interest — a user-defined shape (rectangle, circle or polygon) drawn on the video frame, used to quantify time spent, entries, movement and directionality within specific areas. SHAP SHapley Additive exPlanations — a model-interpretability method giving each feature a contribution score, used in SimBA to explain *why* a :term:`classifier` made a prediction. SLEAP An open-source multi-animal :term:`pose estimation` framework whose output SimBA can import. smoothing Reducing frame-to-frame jitter in tracking data (e.g. Savitzky–Golay or Gaussian) to stabilise body-part trajectories. video info The per-project table (``video_info.csv``) mapping each video to its :term:`FPS`, resolution and :term:`px/mm`. random forest The default supervised algorithm behind a SimBA :term:`classifier`: an ensemble of decision trees whose votes give a per-frame :term:`behavior` probability. cross-validation Splitting annotated data into train/test folds to estimate how well a :term:`classifier` generalises to unseen frames, guarding against over-fitting. feature importance A ranking of how much each :term:`feature` contributes to a :term:`classifier`'s decisions (e.g. Gini importance, permutation importance, or :term:`SHAP`). precision recall F1 Standard classification metrics. Precision = fraction of predicted-positive frames that are correct; recall = fraction of true behavior frames detected; F1 = their harmonic mean. confusion matrix A table of predicted vs. true labels (true/false positives and negatives) used to evaluate a :term:`classifier`. discrimination threshold probability threshold The probability cut-off above which a frame is scored as the :term:`behavior`. Raising it makes detection stricter (higher :term:`precision`), lowering it more permissive (higher :term:`recall`). minimum bout length The shortest allowed :term:`bout` duration; shorter detected episodes are removed as noise during post-classification smoothing. ethogram A catalogue of the distinct behaviors an animal performs, and (in a session) their occurrence over time. keypoint Synonym for :term:`body-part` — a tracked point produced by :term:`pose estimation`. p pose confidence The probability/likelihood score (0–1) that :term:`pose estimation` assigns to each tracked :term:`body-part`, indicating tracking reliability. occlusion When a :term:`body-part` is hidden (by another animal, an object or self) and so is poorly tracked or missing — often handled by :term:`interpolation`. multi-animal tracking identity Tracking several animals at once while maintaining each individual's identity across frames (and recovering it after :term:`occlusion`), e.g. via :term:`maDLC` or :term:`SLEAP`. YOLO A fast real-time object/keypoint detection model family; SimBA supports YOLO-based detection and :term:`pose estimation` workflows. geometry Representing animals/arenas as shapes (points, lines, :term:`convex hull`\ s, polygons, circles) via Shapely, enabling area, overlap, distance and containment computations. centroid The geometric centre of a set of :term:`body-parts ` (or a shape); often used as a single location for an animal. velocity An animal's speed of movement (distance per unit time), typically derived from the frame-to-frame displacement of a :term:`body-part` or :term:`centroid`, in :term:`px/mm`-scaled units. sliding window rolling window A fixed-length time window slid across the data to compute time-resolved :term:`features ` (e.g. mean velocity over the last 0.5 s). time bins Dividing a session into fixed-duration intervals (e.g. 60 s) to report how metrics change over the course of a recording. sequential analysis Analysing the order and timing of behaviors — which tend to precede or follow others (see :term:`FSTTC`) — to uncover behavioral structure. severity scoring Grading the intensity of a detected :term:`behavior` (e.g. attack severity) using movement/feature-based criteria. path plot A visualization tracing an animal's movement trajectory through the arena over time. validation Checking a trained :term:`classifier` on a held-out or new video — including the one-click "validation video" with the predicted probability overlaid frame-by-frame. aggregate statistics Session- or video-level summaries of classifier output (total time, :term:`bout` counts, mean bout duration, latency, etc.) saved to the project ``logs``. clustering embedding Unsupervised grouping of behavioral data without labels — e.g. projecting :term:`features ` with UMAP/t-SNE and clustering the result to discover behavioral motifs.