The outline of CNNs
- Spatial locality & invariance
- Convolution and filters
- Max Pooling
- Example architecture
- Illustrations: what do CNNs not learn?
Chi xinning
小白特别白
MIT+FinTech : ML in Credit Risk Prediction
Neural Network
activiation funciton: typically simple non-linear function to do non-linear transformation.
SWISH function: is continous compared to ReLu.
logistic model 本质是线性的。
Bring in the Economics
How make the engineering solutions speak to the Economics?
more >>MIT+ML07 Neural Network And Optimization
- How to train a NN ?
- What is the basic idea behind training?
- Some of the apsects of NN optimization:
- BP
- Mini-batch
- Initialization
- Batch normalization
- Gradient Clipping
- Adaptive methods: Adam & Adagrad
- Momentum
- Other ideas.
MIT+FinTech Machine Learning in Credit Risk Prediction
ML in Credit Risk Modeling
This is not a typical ML class.
Use the credit Risk model as a lab.
How to apply the tool in a wiser way.
more >>MIT+Fintech Modeling Dynamic User Interests: A Neural Matrix Factorization Approach
Modeling Dynamic User Interests: A Neural Matrix Factorization Approach
Research Context:Boston Globe
- What they really care about?
- Content Recommendation:想变得有趣然后留住customer
- Dynamic User Segementation:
- 什么是Dynamic User Segementation？
- Person A :sports/ Person B:politics
- create profiles, and use this profiles for targeting ads, offers.
- Company want to figure out your interest and target you and hope you buy their product.
- Optimize user experience: how to place articles at the front page?
Our broad contribution : A novel neural network for modeling dynamic of user interests
Desired Output: A sample user’s meaningful content consumption journey
Indentify these latent user interests
more >>MIT+ML06 Intro to Neural Networks
Intro to Neural Networks
Overall Process:
Data-> Feature Extraction -> Classifier
Neural Networks:直接从数据中获取特征进行学习，无需进行特征提取。
That is “turn a feature extraction by hand into a ML task“
简单的分类问题：
get a linear model: h(x)=wTx+b
get a linear model + sigmoid activation: h(x)=σ(wTx+b)
set threshold to do classifier :
ypredicted={0ifh(x)<0.5\1ifh(x)≥0.5
在这里，如果是多分类问题呢，我们就去用softmax。
MIT+ Fintech Use of Machine Learning Methods in Personalization and Targeting
Use of Machine Learning Methods in Personalization and Targeting
- When the methods are useful? Robustness
- Method adaptations: Dynamic Targeting
- Social impact. :Algorithems bias
手动计算深度学习模型中的参数数量
手动计算深度学习模型中的参数数量
我们为什么需要计算DL模型中的参数数量？
More deserve to learn about Efficient Methods and Hardware for DL.
本来呢，代码也可以帮你计算你在模型中所使用的这些trainable parameters of DL models.但是在初期知道如何去计算这些参数对于理解非常有用。
关于以下两类神经网络中的模型参数的计算。
- FFNN
- CNN
more >>Reference: