TaRDIS-APIs/Artificial-Intelligence-and-Machine-Learning-APIs.md

@@ -104,7 +104,7 @@ This API offers a selection of FL algorithms implemented in the Flower framework
 
  The current implementations include:
 
-- **FedAvg**: The well-known federated Averaging algorithm [1], described in D5.1 [3]. The implementation has been also complemented with knowledge distillation, as a result of collaboration within TaRDIS ML teams.
+- **FedAvg**: The well-known federated Averaging algorithm [1], described in D5.1 [3]. The implementation has been also complemented with knowledge distillation, as a result of collaboration between TaRDIS ML teams.
 - **pFedMe - Personalized Federated Learning with Moreau Envelopes**: For personalized model training [2], also described in D5.1 [3]. The algorithm performsa defined number of global and local rounds. Inside the local rounds, the algorithm takes a batch and perform a defined number of inner iterations on it.
 - **pFedMeNew - New personalized Federated Learning with Moreau Envelopes**: A variation of pFedMe, that differs in local iterations and batch sampling. Inside a local round, this algorithm takes a new batch inside every inner iteration.
 - **Anomaly Detection with Noisy Labels**: It integrates two distinct yet complementary machine learning methodologies: clustering using the K-means algorithm and representational learning through autoencoders. This is inspired by [4]. These methods are combined under the name HUNOD - Hybrid UNsupervised Outlier Detection method.