# DETECTING STATISTICAL INTERACTIONS FROM NEURAL NETWORK WEIGHTS

## Contents

# Introduction

It has been commonly believed that one major advantage of neural networks is their capability of modeling complex statistical interactions between features for automatic feature learning. Statistical interactions capture important information on where features often have joint effects with other features on predicting an outcome. The discovery of interactions is especially useful for scientific discoveries and hypothesis validation. For example, physicists may be interested in understanding what joint factors provide evidence for new elementary particles; doctors may want to know what interactions are accounted for in risk prediction models, to compare against known interactions from existing medical literature.

With the growth in the computational power available Neural Networks have been able to solve many of the complex tasks in a wide variety of fields. This is mainly due to their ability to model complex and non-linear interactions. Neural networks have traditionally been treated as “black box” models, preventing their adoption in many application domains, such as those where explainability is desirable. It has been noted that complex machine learning models can learn unintended patterns from data, raising significant risks to stakeholders [14]. Therefore, in applications where machine learning models are intended for making critical decisions, such as healthcare or finance, it is paramount to understand how they make predictions [9]. Within several areas, like eg: computation social science, interpretability is of utmost importance. Since we do not understand how a neural network comes to its decision, practitioners in these areas tend to prefer simpler models like linear regression, decision trees, etc. which are much more interpretable. In this paper, we are going to present one way of implementing interpretability in a neural network.

Existing approaches to interpreting neural networks can be summarized into two types. One type is direct interpretation, which focuses on 1) explaining individual feature importance, for example by computing input gradients [13] and decomposing predictions [8], 2) developing attention-based models, which illustrate where neural networks focus during inference [11], and 3) providing model-specific visualizations, such as feature map and gate activation visualizations [15]. The other type is indirect interpretation, for example post-hoc interpretations of feature importance [12] and knowledge distillation to simpler interpretable models [10].

In this paper, the authors propose Neural Interaction Detection (NID), which can detect any order or form of statistical interaction captured by the feedforward neural network by examining its weight matrix

Note that in this paper, we only consider one specific types of neural network, Feed-Forward Neural Network. Based on the methodology discussed here, the authors suggest that we can build an interpretation methodology for other types of networks also.

# Related Work

1. Interaction Detection approaches:

- Conduct individual tests for all features' combination such as ANOVA and Additive Groves.
- Define all interaction forms of interest, then later finds the important ones.

- The paper's goal is to detect interactions without compromising the functional forms. Our method accomplishes higher-order interaction detection, which has the benefit of avoiding a high false positive or false discovery rate.

2. Interpretability: A lot of work has also been done in this particular area and it can be divided it the following broad categories:

- Feature Importance through Decomposition: Methods like Input Gradient(Sundararajan et al., 2017) learns the importance of features through a gradient-based approach similar to backpropagation. Works like Li et al(2017), Murdoch(2017) and Murdoch(2018) study interpretability of LSTMs by looking at phrase and word level importance scores. Bach et al. 2015 and Shrikumar et al. 2016 (DeepLift) study pixel importance in CNNs.
- Studying Visualizations in Models - Karpathy et al. (2015) worked with character generating LSTMs and tried to study activation and firing in certain hidden units for meaningful attributes. (Yosinski et al., 2015 studies feature map visualizations.
- Attention-Based Models: Bahdanau et al. (2014) - These are a different class of models which use attention modules(different architectures) to help focus the neural network to decide the parts of the input that it should look more closely or give more importance to. Looking at the results of these type of model an indirect sense of interpretability can be gauged.

The approach in this paper is to extract non-additive interactions between variables from the neural network weights.

# Notations

Before we dive in to methodology, we are going to define a few notations here. Most of them will be trivial.

1. Vector: Vectors are defined with bold-lowercases, **v, w**

2. Matrix: Matrice are defined with blod-uppercases, **V, W**

3. Interger Set: For some interger p [math]\in[/math] Z, we define [p] := {1,2,3,...,p}

# Interaction

First of all, in order to explain the model, we need to be able to explain the interactions and their effects to output. Therefore, we define 'interacion' between variables as below.

From the definition above, for a function like, [math]x_1x_2 + sin(x_3 + x_4 + x_5)[/math], we have [math]{[x_1, x_2]}[/math] and [math]{[x_3, x_4, x_5]}[/math] interactions. And we say that the latter interaction to be 3-way interaction.

Note that from the definition above, we can naturally deduce that d-way interaction can exist if and only if all of its (d-1) interactions exist. For example, 3-way interaction above shows that we have 2-way interactions [math]{[3,4], [4,5]}[/math] and [math]{[3,5]}[/math].

One thing that we need to keep in mind is that for models like neural network, most of interactions are happening within hidden layers. This means that we needa proper way of measuring interaction strength.

The key observation is that for any kinds of interaction, at a some hidden unit of some hidden layer, two interacting features the ancestors. In graph-theoretical language, interaction map can be viewed as an associated directed graph and for any interaction [math]\Gamma \in [p][/math], there exists at least one vertix that has all of features of [math]\Gamma[/math] as ancestors. The statement can be rigorized as the following:

Now, the above mathematical statement gurantees us to measure interaction strengths at ANY hidden layers. For example, if we want to study about interactions at some specific hidden layer, now we now that there exists corresponding vertices between the hidden layer and output layer. Therefore all we need to do is now to find approprite measure which can summarize the information between those two layers. } Before doing so, let's think about a single-layered neural network. For any one hidden unit, we can have possibly, [math]2^{||W_i,:||}[/math], number of interactions. This means that our search space might be too huge for multi-layered networks. Therefore, we need a some descent way of approximate out search space. Moreover, the authors realized a fast interaction detection by limiting the search complexity of the task by only quantifying interactions created at the first hidden layer.

As we discussed above, in order to consider interaction between units in any layers, we need to think about their out-going paths. However, we soon encountered the fact that for some fully-connected multi-layer neural network, the search space might be too huge to compare. Therefore, we use information about out-going paths gredient upper bond. To represent the influence of out-going paths at [math]l[/math]-hidden layer, we define cumulative impact of weights between output layer and [math]l+1[/math]. We define aggregated weights as,

Note that [math]z^{(l)} \in R^{(p_l)}[/math] where [math]p_l[/math] is the number of hidden units in [math]l[/math]-layer.
Moreover, this is the lipschitz constant of gredients. Gredient has been an import variable of measuring influence of features, especially when we consider that input layer's derivative computes the direction normal to decision boundaries.

## Quantifying influence

For some [math]i[/math] hidden unit at the first hidden layer, which is the closet layer to the input layer, we define the influence strength of some interaction as,

The function [math]\mu[/math] will be defined later. Essentially, the formula shows that the strength of influence is defined as the product of the aggregated weight on the first hidden layer and some measure of influence between the first hidden layer and the input layer.

For the function, [math]\mu[/math], any positive-real valued functions such as max, min and average can be candidates. The effects of those candidates will be tested later.

Now based on the specifications above, the author suggested the algorithm for searching influential interactions between input layer units as follows:

# Cut off Model

Now using the greedy algorithm defined above, we can rank the interactions by their strength. However, in order to access true interactions, we are building the cut off model which is a generalized additive model (GAM) as below,

From the above model, each [math]g[/math] and [math]g^*[/math] are Feed-Forward neural network. We are keep adding interactions until the performance reaches plateaus.

# Experiment

For the experiment, we are going to compare three neural network model with traditional statistical interaction detecting algorithms. For the nueral network models, first model will be MLP, second model will be MLP-M, which is MLP with additional univariate network at the output. The last one is the cut-off model defined above, which is denoted by MLP-cutoff. MLP-M model is graphically represented below.

For the experiment, We study our interaction detection framework on both simulated and real-world experiments. For simulated experiments, we are going to test on 10 synthetic functions as shown in table I.

We use four real-world datasets, of which two are regression datasets, and the other two are binary classification datasets. The datasets are a mixture of common prediction tasks in the cal housing and bike sharing datasets, a scientific discovery task in the higgs boson dataset, and an example of very-high order interaction detection in the letter dataset.

And the author also reported the results of comparisons between the models. As you can see, neural network based models are performing better in average. Compare to the traditional methods liek ANOVA, MLP and MLP-M method shows 20% increases in performance.

The above result shows that MLP-M almost perfectly catch the most influential pair-wise interactions.

# Limitations

Even though for the above synthetic experiment MLP methods showed superior performances, the method still have some limitations. For example, fir the function like, [math]x_1x_2 + x_2x_3 + x_1x_3[/math], neural network fails to distinguish between interlinked interactions to single higher order interaction. Moreoever, correlation between features deteriorates the ability of the network to distinguish interactions. However, correlation issues are presented most of interaction detection algorithms.

Because this method relies on the neural network fitting the data well, there are some additional concerns. Notably, if the NN is unable to make an appropriate fit (under/overfitting), the resulting interactions will be flawed. This can occur if the datasets that are too small or too noisy, which often occurs in practical settings.

# Conclusion

Here we presented the method of detecting interactions using MLP. Compared to other state-of-the-art methods like Additive Groves (AG), the performances are competitive yet computational powers required is far less. Therefore, it is safe to claim that the method will be extremly useful for practitioners with (comparably) less computational powers. Moreover, the NIP algorithm successfully reduced the computation sizes. After all, the most important aspect of this algorithm is that now users of nueral networks can impose interpretability in the model usage, which will change the level of usability to another level for most of practitioners outside of those working in machine learning and deep learning areas. For future work, the authors want to detect feature interactions by using the common units in the intermediate hidden layers of feedforward networks, and also want to use such interaction detection to interpret weights in other deep neural networks.

# Critique

1. Authors need to do large-scale experiments, instead of just conducting experiments on some synthetic dataset with small feature dimensionality, to make their claim stronger.

2. Although the method proposed in this paper is interesting, the paper would benefit from providing some more explanations to support its idea and fill the possible gaps in its experimental evaluation. In some parts there are repetitive explanations that could be replaced by other essential clarifications.

# Reference

[1] Jacob Bien, Jonathan Taylor, and Robert Tibshirani. A lasso for hierarchical interactions. Annals of statistics, 41(3):1111, 2013.

[2] G David Garson. Interpreting neural-network connection weights. AI Expert, 6(4):46–51, 1991.

[3] Yotam Hechtlinger. Interpretation of prediction models using the input gradient. arXiv preprint arXiv:1611.07634, 2016.

[4] Shiyu Liang and R Srikant. Why deep neural networks for function approximation? 2016.

[5] David Rolnick and Max Tegmark. The power of deeper networks for expressing natural functions. International Conference on Learning Representations, 2018.

[6] Daria Sorokina, Rich Caruana, and Mirek Riedewald. Additive groves of regression trees. Machine Learning: ECML 2007, pp. 323–334, 2007.

[7] Simon Wood. Generalized additive models: an introduction with R. CRC press, 2006

[8] Sebastian Bach, Alexander Binder, Gre ́goire Montavon, Frederick Klauschen, Klaus-Robert Mu ̈ller, and Wojciech Samek. On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PloS one, 10(7):e0130140, 2015.

[9] Rich Caruana, Yin Lou, Johannes Gehrke, Paul Koch, Marc Sturm, and Noemie Elhadad. Intel- ligible models for healthcare: Predicting pneumonia risk and hospital 30-day readmission. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1721–1730. ACM, 2015.

[10] Zhengping Che, Sanjay Purushotham, Robinder Khemani, and Yan Liu. Interpretable deep models for icu outcome prediction. In AMIA Annual Symposium Proceedings, volume 2016, pp. 371. American Medical Informatics Association, 2016.

[11] Laurent Itti, Christof Koch, and Ernst Niebur. A model of saliency-based visual attention for rapid scene analysis. IEEE Transactions on pattern analysis and machine intelligence, 20(11):1254– 1259, 1998.

[12] Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. Why should i trust you?: Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1135–1144. ACM, 2016.

[13]Karen Simonyan, Andrea Vedaldi, and Andrew Zisserman. Deep inside convolutional networks: Vi- sualising image classification models and saliency maps. arXiv preprint arXiv:1312.6034, 2013.

[14] Kush R Varshney and Homa Alemzadeh. On the safety of machine learning: Cyber-physical sys- tems, decision sciences, and data products. arXiv preprint arXiv:1610.01256, 2016.

[15] Jason Yosinski, Jeff Clune, Anh Nguyen, Thomas Fuchs, and Hod Lipson. Understanding neural networks through deep visualization. arXiv preprint arXiv:1506.06579, 2015.