docs: 添加题目名称技术路线预期结果文档

- 新增毕业设计题目说明、技术路线规划和预期结果描述
   - 优化深度学习模型代码，支持 PyTorch 可选依赖

backend/core/deep_learning_model.py

@@ -1,3 +1,4 @@
+from __future__ import annotations
 import copy
 import os
 from typing import Dict, List, Optional, Tuple
@@ -58,6 +59,9 @@ DEFAULT_BATCH_SIZE = 256
 EARLY_STOPPING_PATIENCE = 12
 
 
+BaseTorchModule = nn.Module if nn is not None else object
+
+
 class SequenceStaticDataset(Dataset):
     def __init__(
         self,
@@ -86,7 +90,7 @@ class SequenceStaticDataset(Dataset):
         )
 
 
-class LSTMMLPRegressor(nn.Module):
+class LSTMMLPRegressor(BaseTorchModule):
     def __init__(
         self,
         seq_num_dim: int,