The Neural Tangent Kernel (NTK) has become a powerful framework for analyzing the behavior of deep neural networks in the infinite-width limit. This paper presents a concise overview of the key theoretical foundations of NTK, covering its origins, the proof of deterministic behavior at initialization, and its role in bounding the training loss for regression tasks. Additionally, we extend this analysis by establishing a bound for the training loss in classification problems. Each theoretical property of the NTK is validated through experiments on various datasets.