Datasets & DataLoaders PyTorch provides two data primitives (基本資料型態) that allow you to use pre-loaded datasets as well as your own data, as below: torch.utils.data.Dataset Stores the samples and their corresponding labels Dataset 定義資料的結構並且將其包起來，利如: 一張影像和一個標籤 一張影
Tensors Tensors are a specialized data structure that are very similar to arrays and matrices. In PyTorch, we use tensors to encode the inputs and outputs of a model, as well as the model’s parameters. 張量 (Tensor) 類似向量或矩陣，他是一個 n-維 (n-dimensional) 的資料型態，像是: 0
目標檢測評估指標 (Object Detection Metrics) 2024/6/4 updated! Object Detection 需要執行兩個任務: decide whether an object exists in the scene: classification determine the position, the orientation and the shape of the object: localization 在現實場景中，一個場景會包含很多物件類別（如: 車輛
分類指標 (Classification Metrics) 2024/6/4 updated! Confuion Matrix (混淆矩陣) 分類器的目的在對樣本進行分類，例如，將樣本中的男女進行區分。不過，在預測分類的過程中，會有預測正確與預測錯誤
Matrix Chain Multiplication Input: a sequence of n matries $(A_1, …, A_n)$ Output: the product of $A_1A_2 … A_n$ 根據 Associative (結合律) 可以將問題重新定義為: Matrix Chain Multiplication Problem 相關說明可參考: Observation Observation Each entry takes $q$ multiplications There are total $pr$ entries Time-Complexity: $\Theta(q)\Theta(pr) = \Theta(pqr)$ Associative (結合