**Augmented Intelligence – An Elegant Approach to Optimize Your Decision Making Process**

Automation is a rising trend in the recent technology boom, but it can impose a level or risk. Harmonizing both human decision and powerful computing abilities will be key, especially for enterprises looking to unlock insights through analytics.

**H2O World – Top 10 Deep Learning Tips & Tricks – Arno Cand**

1. Understand Model Complexity

• Model size depends on features

• Model size is independent of number of rows or training time

2. Establish a Baseline on Holdout Data

• Develop a feel for the problem and the holdout performance of the different models

3. Inspect Models in Flow (Notebook-style open-source UI for H2O)

• If model is wrong(wrong architecture, response, parameters, etc.), cancel it

• If model is taking too long, cancel and decrease model complexity

• If model is performing badly, cancel and increase model complexity

4. Use Early Stopping (On by default for H2O)

• Saves tons of time

• Use Flow to inspect model

5. Control Scoring Overhead

• Validation data determines scoring speed and early stopping decision

6. Use N-fold Cross-Validation

• Estimate your model performance well

7. Use Regularization

• Overfitting is easy, generalization is art

8. Perform HyperParameter Search

• Parameters to tune: hidden dropout ratios, input dropout ratios, adaptive rate, etc.

• Focus on just finding one of the many good models

9. Use Checkpointing

• Checkpointing enables fast exploration

10. Tune Communication on Multi-Node

• Model size depends on features

• Model size is independent of number of rows or training time

2. Establish a Baseline on Holdout Data

• Develop a feel for the problem and the holdout performance of the different models

3. Inspect Models in Flow (Notebook-style open-source UI for H2O)

• If model is wrong(wrong architecture, response, parameters, etc.), cancel it

• If model is taking too long, cancel and decrease model complexity

• If model is performing badly, cancel and increase model complexity

4. Use Early Stopping (On by default for H2O)

• Saves tons of time

• Use Flow to inspect model

5. Control Scoring Overhead

• Validation data determines scoring speed and early stopping decision

6. Use N-fold Cross-Validation

• Estimate your model performance well

7. Use Regularization

• Overfitting is easy, generalization is art

8. Perform HyperParameter Search

• Parameters to tune: hidden dropout ratios, input dropout ratios, adaptive rate, etc.

• Focus on just finding one of the many good models

9. Use Checkpointing

• Checkpointing enables fast exploration

10. Tune Communication on Multi-Node

**R and Python: Theory of Linear Least Squares**

In my previous article, we talked about implementations of linear regression models in R, Python and SAS. On the theoretical sides, however, I briefly mentioned the estimation procedure for the parameter β . So to help us understand how software does the estimation procedure, we’ll look at the mathematics behind it. We will also perform the estimation manually in R and in Python, that means we’re not gonna use any special packages, this will help us appreciate the theory.

**10 Machine Learning Algorithms Explained to an ‘Army Soldier’**

In this article, I’ve revealed that approach used to explain ML algorithms to the soldier. Though, there would be less soldiers who would read this, yet this idea promotes an interesting way of learning. This would surely help beginners who are struggling to understand these algorithms. I’ve also added images to help you visualize the situation and learn.

**Sampling Nearly-Equidistant Points**

I ran into a curious problem this week. Let’s say you have unevenly-spaced data, and you want to pick the subset that is most nearly equally spaced. How do you do it? I found a stackoverflow post that gave a really cool answer, but since the post didn’t come with a lot of explanation, I thought I’d talk about it a bit.

**Breaking into the Indian E-commerce**

Myntra is India’s largest online fashion and lifestyle store, as they say. Which is owned by Flipkart, one of the big players in Indian E-commerce. I found a little opening in the flow of their online movement and used it to collect some information(not private though). Using this information we can see the way ‘online shopping’ is escalating in the country, and the role Myntra and the people are playing in it. Sounds interesting? Hold on.

**Top 10 Machine Learning Projects on Github**

1. Scikit-learn

2. Awesome Machine Learning

3. PredictionIO

4. Dive Into Machine Learning

5. Pattern

6. NuPIC (Numenta Platform for Intelligent Computing)

7. Vowpal Wabbit

8. aerosolve

9. GoLearn

10. Code for Machine Learning for Hackers

2. Awesome Machine Learning

3. PredictionIO

4. Dive Into Machine Learning

5. Pattern

6. NuPIC (Numenta Platform for Intelligent Computing)

7. Vowpal Wabbit

8. aerosolve

9. GoLearn

10. Code for Machine Learning for Hackers

**Graphs: when zero is not a relevant value**

Bar charts have a filled area tying the axis to the plotted value, and this only makes sense when the axis is at a true zero. Scatterplots and line plot don’t have the same limitations, and can be useful even when there isn’t a true zero or it isn’t a relevant value.

**Importance of Data Science for IoT business**

Here, we have explored how IoT businesses can leverage data science for IT strategies, service analysis stack, capacity planning, hardware maintenance, competitive advantages and anomaly detection. Along with, the different application in multiple IoT in

**Making Sense of Logarithmic Loss**

Logarithmic Loss, or simply Log Loss, is a classification loss function often used as an evaluation metric in kaggle competitions. Since success in these competitions hinges on effectively minimising the Log Loss, it makes sense to have some understanding of how this metric is calculated and how it should be interpreted.