fix run.sh and add readme

+## 상관계수 정규화와 동적 필터 가지치기를 이용한 심층 신경망 압축
+Dynamic Filter Pruning with Decorrelation Regularization for Compression of Deep Neural Network
+> 2020 한국소프트웨어종합학술대회 (KSC2020) 학부생논문 경진대회 출품작
+
+### Prerequisites
+
+* Ubuntu 18.04
+* Python 3.7.4
+* Pytorch 1.6.0
+* numpy 1.18.1
+* GPU (cuda)
+
+### Build
+
+```
+$ python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.7 > result.txt
+```
+* `run.sh`에서 parameter 조절 후 `./run.sh`로 진행
+
+### Process
+**0. Data, Model & Parameters**
+- Data : CIFAR-10
+- Model : ResNet-50
+- Optimizer : Stochastic Gradient Descent
+- Learning Rate : 0.2
+- Epoch : 300
+- Batch size : 128
+- Loss Function : Cross Entropy
+- Metric : Accuracy, Sparsity
+
+**1. 동적 필터 가지치기 (Dynamic Filter Pruning)**
+
+L1 norm 크기를 기반으로 필터 마스크를 생성하여 가중치 학습 시 반영
+- 필터 마스크 : ![image](https://user-images.githubusercontent.com/41580746/102396051-41616d80-401f-11eb-9738-7b5df9aee0d4.png)
+   - i : 층 위치
+   - j : 필터 위치
+   - t : epoch 값
+   - W : 필터 가중치 행렬
+   - η : 임계값 (전체 필터 개수 중 가지치기 필터 개수 비율 통해 계산)
+- 가중치 학습 : ![image](https://user-images.githubusercontent.com/41580746/102396101-51794d00-401f-11eb-9303-99dd712798ee.png)
+   - g : 기울기
+   - γ : learning rate
+
+**2. 상관계수 정규화 (Decorrleation Regularization)**
+
+기존 loss function에 상관계수 정규화 식을 더하여 최종 손실 함수 계산
+- loss function : ![image](https://user-images.githubusercontent.com/41580746/102396482-d82e2a00-401f-11eb-93b3-7db5fea8f8af.png)
+   - α : 정규화 상수
+   - ![image](https://user-images.githubusercontent.com/41580746/102396603-01e75100-4020-11eb-933e-f85305dd874d.png)
+
+### Result
+가지치기 비율 60%, 정규화 상수 0.7일 때의 모델별 Accuracy 및 Sparsity 비교 결과
+- ![image](https://user-images.githubusercontent.com/41580746/102395542-a072b280-401e-11eb-9e47-c3b52d859479.png)
+- ![image](https://user-images.githubusercontent.com/41580746/102396706-28a58780-4020-11eb-9ebb-6cc723b4fcbf.png)
+- 기존 동적 필터 가지치기 대비 Accuracy 1.47%, Sparsity 1.08% 증가
+
+### References
+- [1] Yann LeCun, Yoshua Bengio, Geoffrey Hinton. Deep learning. Nature 521, 436-444, 2015.
+- [2] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jain Sun. Deep Residual Learning for Image Recognition. 2015.
+- [3] 조인천, 배성호. 동적 필터 프루닝 기법을 이용한 심층 신경망 압축. 한국방송미디어공학회 하계학술대회, 2020.
+- [4] Benoit Jacob. Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference. 2017.
+- [5] Tao Lin, Sebastian U. Stich, Luis Barba, Daniil Dmitriev, Martin Jaggi. Dynamic Model Pruning with Feedback. ICLR, 2020.
+- [6] Namhoon Lee, Thalaiyasingam Ajanthan, Philip HS Torr, SNIP: Single-shot network pruningbased on connection sensitivity. ICLR, 2019.
+- [7] Hao Li, Asim Kadav, Igor Durdanovic, Hanan Samet, Hans Peter Graf. Pruning Filters For Effiecient ConvNets. ICLR, 2017.
+- [8] Jian-Hao Luo, Jianxin Wu, Weiyao Lin. ThiNet: A Filter Level Pruning Method for Deep Neural Network Compression. ICCV, 2017.
+- [9] Song Han, Huizi Mao, William J. Dally. Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding. ICLR, 2016.
+- [10] Xiaotian Zhu, Wengang Zhou, Houqiang Li. Improving Deep Neural Network Sparsity through Decorrelation Regularization. IJCAI, 2018.

 #!/bin/bash
-RESULT_DIR=result_201203
+RESULT_DIR=result
 
 if [ ! -d $RESULT_DIR ]; then
     mkdir $RESULT_DIR
 fi
 
-#python modeling_default.py > $RESULT_DIR/default.txt #&
-#python modeling_pruning.py > $RESULT_DIR/pruning_prune90.txt &
-#python modeling_decorrelation.py > $RESULT_DIR/decorrelation_lambda1.txt &
-#python modeling_pruning+decorrelation.py > $RESULT_DIR/pruning+decorrelation_lambda1+prune90.txt
 
-#python modeling.py --prune_type structured --prune_rate 0.5 > $RESULT_DIR/prune_05.txt
-#python modeling.py --prune_type structured --prune_rate 0.6 > $RESULT_DIR/prune_06.txt
-#python modeling.py --prune_type structured --prune_rate 0.8 > $RESULT_DIR/prune_08.txt &
-#python modeling.py --prune_type structured --prune_rate 0.7 > $RESULT_DIR/prune_07.txt
+#python modeling.py --prune_type structured --prune_rate 0.5 > $RESULT_DIR/prune_50.txt
+#python modeling.py --prune_type structured --prune_rate 0.6 > $RESULT_DIR/prune_60.txt
+#python modeling.py --prune_type structured --prune_rate 0.7 > $RESULT_DIR/prune_70.txt
+#python modeling.py --prune_type structured --prune_rate 0.8 > $RESULT_DIR/prune_80.txt
 
-#python modeling.py --reg reg_cov --odecay 0.9 > $RESULT_DIR/reg_9.txt
-#python modeling.py --reg reg_cov --odecay 0.8 > $RESULT_DIR/reg_8.txt
-#python modeling.py --reg reg_cov --odecay 0.7 > $RESULT_DIR/reg_7.txt
-#python modeling.py --reg reg_cov --odecay 0.6 > $RESULT_DIR/reg_6.txt
-#python modeling.py --reg reg_cov --odecay 0.5 > $RESULT_DIR/reg_5.txt
+#python modeling.py --reg reg_cov --odecay 0.5 > $RESULT_DIR/reg_05.txt
+#python modeling.py --reg reg_cov --odecay 0.6 > $RESULT_DIR/reg_06.txt
+#python modeling.py --reg reg_cov --odecay 0.7 > $RESULT_DIR/reg_07.txt
+#python modeling.py --reg reg_cov --odecay 0.8 > $RESULT_DIR/reg_08.txt
+#python modeling.py --reg reg_cov --odecay 0.9 > $RESULT_DIR/reg_09.txt
 
-#python modeling.py --prune_type structured --prune_rate 0.5 --reg reg_cov --odecay 0.7 > $RESULT_DIR/prune_05_reg_07.txt &
-#python modeling.py --prune_type structured --prune_rate 0.5 --reg reg_cov --odecay 0.8 > $RESULT_DIR/prune_05_reg_08.txt &
-#python modeling.py --prune_type structured --prune_rate 0.5 --reg reg_cov --odecay 0.9 > $RESULT_DIR/prune_05_reg_09.txt &
-#python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.7 > $RESULT_DIR/prune_06_reg_07.txt &
-#python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.8 > $RESULT_DIR/prune_06_reg_08.txt &
-python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.9 > $RESULT_DIR/prune_06_reg_09.txt 
 
+#python modeling.py --prune_type structured --prune_rate 0.5 --reg reg_cov --odecay 0.7 > $RESULT_DIR/prune_50_reg_07.txt
+#python modeling.py --prune_type structured --prune_rate 0.5 --reg reg_cov --odecay 0.8 > $RESULT_DIR/prune_50_reg_08.txt
+#python modeling.py --prune_type structured --prune_rate 0.5 --reg reg_cov --odecay 0.9 > $RESULT_DIR/prune_50_reg_09.txt
+python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.7 > $RESULT_DIR/prune_60_reg_07.txt
+#python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.8 > $RESULT_DIR/prune_60_reg_08.txt
+#python modeling.py --prune_type structured --prune_rate 0.6 --reg reg_cov --odecay 0.9 > $RESULT_DIR/prune_60_reg_09.txt