cosmetic changes

Author: ryanrhymes

chapters/algodiff.md

@@ -1560,5 +1560,3 @@ You can see that the core idea of AD can be implemented with surprisingly simple
 Then we turn to the Owl side: first, how Owl support what we have done in the strawman implementation with the forward and reverse propagation APIs; next, how Owl provides various powerful high level APIs to enable users to directly perform AD.
 Finally, we give an in-depth introduction to the implementation of the AD module in Owl, including some details that enhance the simple strawman code, how to build user-defined AD computation, and using lazy evaluation to improve performance, etc.
 Hopefully, after finishing this chapter, you can have a solid understanding of both its theory and implementation.
-
-## References

chapters/diffequation.md

@@ -831,5 +831,3 @@ Also, this library interfaces to existing off-the-shelf open source ODE solvers
 Next, we demonstrate how these solvers are used to solve ODE from several real examples, including the Two-body problem, the Lorentz Attractor, and Damped Oscillation, etc.
 Finally, we introduce one important idea in the solving ODE numerically: stiffness, and then shows how we can solve stiff and non-stiff ODEs with the example of the van der pol equation.
 Hopefully, after studying this chapter, the readers can have a basic idea of how numerical ODE solver works and how to apply them into solving real-world problems.
-
-## References

chapters/linalg.md

@@ -1452,4 +1452,3 @@ Sure enough, there is no way to cover all these topics in one chapter. We refer
 In the end, we introduce how the linear algebra module is implemented in a numerical library such as Owl. 
 We close the discussion with a brief explanation of the sparse matrix and the representation formats used.
 
-## References

chapters/ndarray.md

@@ -700,4 +700,3 @@ This chapter explain in detail the Ndarray module, including its creation, prope
 Besides, we also discuss the subtle difference between tensor and ndarray in this chapter.
 This chapter is easy to follow, and can serve as a reference whenever users need a quick check of functions they need.
 
-## References

chapters/neural-network.md

@@ -917,4 +917,3 @@ Here we use the Owl API to solve the same example, including training and testin
 We then introduce two important types of neural network: the convolutional neural network, together its superior performance against simple feedforward network, and the recurrent neural network, including two of its variants: the LSTM and GRU.
 We finish this chapter with a brief introduction of the basic idea behind Generative Adversarial Network, another type of neural network that has gained a lot of momentum in research and application recently.
 
-## References

chapters/optimisation.md

@@ -517,4 +517,3 @@ The local unconstrained problem is further explained in two parts: the univariat
 Of all the methods introduced here, the gradient descent is especially important, and we will see it again in the Regression and Neural Network chapters.
 Finally, we give a brief peek at the topic of global and constrained  optimisation problems.
 
-## References

chapters/regression.md

@@ -950,4 +950,3 @@ With a bit of change in its cost function, we venture to introduce a type of adv
 They can be used for both linear and non-linear decision boundary by using different kernel functions.
 We have also talked about related issues, such as the regularisation, model error and selection.
 
-## References

chapters/signal.md

@@ -739,4 +739,3 @@ Finally, we discussed filtering in signal process using different techniques, in
 Here we also explained the relationship between the two most crucial computations in numerical applications: FFT and convolution.
 More about convolution can be find in the Neural Network chapter.
 
-## References

chapters/stats.md

@@ -632,4 +632,3 @@ Then we went from descriptive statistics to inference statistics, and introduced
 Next, we covered the basic idea in hypothesis testing with examples. 
 The difference between covariance and correlations is also discussed.
 
-## References

chapters/visualization.md

@@ -683,5 +683,3 @@ The good news is that, you can easily change the colors in the plot.
 Try google "colour theory" and you can find a lot of guidelines. 
 For example, the analogous colours can be a good choice in your line plots. These colours are any sequential three colors on a 12-part color wheel, such as yellow-green, yellow, and yellow-orange.
 We find this artistic aspect of visualisation is often enjoyable.
-
-## References

index.markdown

@@ -32,12 +32,3 @@ The main purpose of the online tutorial is for teaching how to use the [Owl](htt
 
 Currently contribution to the book is mainly in the form of Pull Request on [GitHub](https://github.com/owlbarn/tutorial). 
 Normally you only need to change one of the markdown files in the `chapters/` directory.
-
-
-## Tooling
-
-The following tools are used in the project, please refer to their documentation.
-
-- [Jekyll](https://jekyllrb.com/)
-- [Cayman Theme](https://github.com/pages-themes/cayman)
-