One advantage of our algorithm is that it is not binary in nature. by Yong Kuk Kim (13955214)

Key steps in neurodegeneration detection algorithm. by Andrew S. Clark (16510496)

Algoritmo de detección de odio en español (Algorithm for detection of hate speech in Spanish) by Elias Said-Hung (10790310)

“…</li></ul><h2>Training Process</h2><h3>Pre-workout</h3><ul><li>Batch size: 16</li><li>Epochs: 5</li><li>Learning rate: 2e-5 with 10% warmup steps</li><li>Early stopping with patience=2</li></ul><h3>Fine-tuning</h3><ul><li>Batch size: 128</li><li>Epochs: 5</li><li>Learning rate: 2e-5 with 10% warmup steps</li><li>Early stopping with patience=2</li><li>Custom metrics:</li><li>Recall for non-hate class</li><li>Precision for hate class</li><li>F1-score (weighted)</li><li>AUC-PR</li><li>Recall at precision=0.9 (non-hate)</li><li>Precision at recall=0.9 (hate)</li></ul><h2>Evaluation Metrics</h2><p dir="ltr">The model is evaluated using:</p><ul><li>Macro recall, precision, and F1-score</li><li>One-vs-Rest AUC</li><li>Accuracy</li><li>Metrics by class</li><li>Confusion matrix</li></ul><h2>Requirements</h2><p dir="ltr">The following Python packages are required (see requirements.txt for the full list):</p><ul><li>TensorFlow</li><li>Transformers</li><li>scikit-learn</li><li>pandas</li><li>datasets</li><li>matplotlib</li><li>seaborn</li></ul><h2>Usage</h2><p dir="ltr">The model expects input data with the following specifications:</p><ol><li><b>Data Format</b>:</li></ol><ul><li>CSV file or Pandas DataFrame</li><li>Mandatory column name: <code>text</code> (type string)</li><li>Optional column name: <code>label</code> (type integer, 0 or 1) if available for evaluation</li></ul><ol><li><b>Text Preprocessing</b>:</li></ol><ul><li>Text will be automatically converted to lowercase during processing</li><li>Maximum length: 128 tokens (longer texts will be truncated)</li><li>Special characters, URLs, and emojis must remain in the text (the tokenizer handles these)</li></ul><ol><li><b>Label Encoding</b>:</li></ol><ul><li><code>0</code> = No hateful content (including neutral/positive content)</li><li>1 = Hate speech</li></ul><p dir="ltr">The process of creating this algorithm is explained in the technical report located at:Blanco-Valencia, X., De Gregorio-Vicente, O., Ruiz Iniesta, A., & Said-Hung, E. (2025). …”

Particle tracking validation of the S8 algorithm. by Jennifer M. Knighten (15453354)

Data set constituents. by Rémy Peyret (14691736)

“…We describe a patch-based algorithm that incorporates a convolutional neural network to detect and locate invasive carcinoma on breast whole-slide images. …”

Cell classification. by Adam L. Tyson (10892822)

Results of the model on test sets 1 and 2. by Rémy Peyret (14691736)

“…We describe a patch-based algorithm that incorporates a convolutional neural network to detect and locate invasive carcinoma on breast whole-slide images. …”

Scanners and staining methods. by Rémy Peyret (14691736)

“…We describe a patch-based algorithm that incorporates a convolutional neural network to detect and locate invasive carcinoma on breast whole-slide images. …”

Algoritmo de clasificación de expresiones de odio por tipos en español (Algorithm for classifying hate expressions by type in Spanish) by Daniel Pérez Palau (11097348)

“…</li></ul><p dir="ltr"><b>File Structure</b></p><p dir="ltr">The code generates and saves:</p><ul><li>Weights of the trained model (.h5)</li><li>Configured tokenizer</li><li>Training history in CSV</li><li>Requirements file</li></ul><p dir="ltr"><b>Important Notes</b></p><ul><li>The model excludes category 2 during training</li><li>Implements transfer learning from a pre-trained model for binary hate detection</li><li>Includes early stopping callbacks to prevent overfitting</li><li>Uses class weighting to handle category imbalances</li></ul><p dir="ltr">The process of creating this algorithm is explained in the technical report located at: Blanco-Valencia, X., De Gregorio-Vicente, O., Ruiz Iniesta, A., & Said-Hung, E. (2025). …”

Model & data for automated classification of male and female house wren song by Karan Odom (18271399)

Fig 1 - A Discretized Overlap Resolution Algorithm (DORA) for resolving spatial overlaps in individual-based models of microbes by Ihab Hashem (21156806)

Bootstrap-based inference for multiple mean-variance changepoint models by Yang Li (7082)

“…This method integrates the weighted bootstrap with the Sequential Binary Segmentation (SBS) algorithm. Not only does our technique pinpoint the location and number of change points, but it also determines the type of change for each estimated point, specifying whether the change occurred in the mean, variance, or both. …”

Behavioral and Eye-tracking Data for Adaptive Circuit Dynamics Across Human Cortex During Evidence Accumulation in Changing Environments by Peter Murphy (10141523)

“…The main variable in each file is a matrix called <i>Behav</i> for which each row is a trial and columns are the following:</p> <p>column 1 – the generative distribution used to draw the final sample location on each trial (and thus, the correct response)</p> <p>column 2 – the response given by the participant</p> <p>column 3 – the accuracy of the participant’s response</p> <p>column 4 – response time relative to Go cue</p> <p>column 5 – trial onset according to psychtoolbox clock</p> <p>column 6 – number of times participant broke fixation during trial, according to online detection algorithm</p> <p>Each .mat file also contains a trials*samples matrix (<i>tRefresh</i>) of the timings of monitor flips corresponding to the onsets of each sample (and made relative to trial onset), as provided by psychtoolbox.…”