Descriptive Statistics for Data Analytics/Science

Basically descriptive statistics tells everything about the data that is available, it doesn’t tell us anything out of that data.

So how to describe the (Numerical/Categorical) data?

- graphical representation 

- tabular representation

- summary statistics 

Type of variables in statistics-

1) Numerical or Quantitative (continuous and discrete)

2) Categorical or Qualitative (only discrete)

- nominal

- ordinal

Speaking about variables they can be single or multiple as 

We deal with data in descriptive statistics, it is important to establish relationship between this variables so as to summarize.

• Graphical representation- Single variable

- For Categorical(Qualitative) variables:

-> Bar Graph: suited for ordinal variables 

-> Pie Chart: suited for nominal variables

- For Quantitative(Numerical) variables:

-> Box-plot( suited for continuous data)

-> Histogram(1st step to start before distribution)

-suited for discrete data

• Graphical Representation: Multiple variables

- Scatter plots-> two quantitative (numerical) variables

- Box-plots-> one categorical (qualitative) other quantitative variable (Suited for continuous data)

- Contingency table -> both categorical with frequency of occurrence of both.

• Measures of Central Tendency: 

How to summarize data through numbers?(summary statistics)

- Measures of central tendency (mean, mode, median)

- Dispersion

- Skew and Kurtosis

Choosing when to use mean, median:

- bad outlier

- good outlier

When to use Mode:

- useful with nominal variables 

- multi modal distribution

• Measures of dispersion:

- Range (Max- Min)
- InterQuartile Range (IQR) = 3rd Quartile - 1st Quartile or (75th percentile - 25th percentile) 
- Sample Standard Deviation
- Variance
- Mean Absolute Deviation

 All Formulas:

How Amazon Echo works??

The amazon echo is a smart IOT device which listens to your voice and translates it into commands to perform tasks such as Turn ON/OFF lights, playing music, ordering products from amazon, setting alarm, etc. Read this blogpost to know how it works!!!

What is Amazon Echo?

Amazon echo is a device which uses speech recognition to perform tasks, the actual brain which performs these tasks is Alexa. Alexa  uses a small computer along with microphones, speakers, and cloud computing (AWS). It performs a wide range of tasks using the 'Internet'. All you need is smart devices around you to make use of the echo fully. 

How Echo/Alexa works?

Echo is always ON, it connect's to the home wifi and stays ON. When you say something "Alexa..." it wakes up, Alexa records that voice and internally gathers voice commands and send them to the natural langauge voice recognition service in the clouds called Alexa Voice Service it interprets the commands and sends back the result.

Check this graphical representation to know how it works.

 What are Alexa compatible devices?

• Echo
• Echo dot
• Echo show
• Echo spot
• Echo plus
• Echo Input
• Echo Show 5
• Echo Tap

Buy Echo devices from amazon Buy here

What all home devices are compatible with Alexa other than Echo (smart speaker)?

Arlo Q - Smart Camera buy here

Ecobee (5th gen) - smart thermostat buy here

Philips hue white A19 starter kit - smart lights  buy here

Check this post for more updates about amazon Alexa. 

Delete the Cam scanner app now!!!

Your portable mobile scanner is infected with malware!!

We come across many apps in our smart phones which triggers lots of advertisements(almost all free apps). As we tend to use free version of apps, we don’t have privilege to disable the Ads. If you want seamless ad free experience you need to spend bucks for purchasing the premium version of app. Recently cam scanner was surrounded with new controversy which looks very scary‼️‼️

Cam Scanner has been removed from the plays store!!

The malware is Trojan Dropper, it’s a software which runs and downloads additional modules on the smartphone without the user consent. In case of cam-scanner the malware runs multiple ads on the phone  and make user signup for paid services without there knowledge. Now let’s understand in layman terms how critical is this issue!!!

What is Malware?

Malware is a program/file that is harmful to a computer/smartphone user. Malware can be viruses, worms, Trojan horses or spyware. This program performs different types of functions like stealing user data, taking control over a specific computer function or system. Malware can monitor or even delete sensitive data without user's permission.

What is Cam scanner?

It's a widely used mobile app which converts your phone into a portable scanner. It helps to scan documents, ID's, images etc. It automatically crops the edges and enhance the scanned document. It has more than 100 million downloads on play store.

Why should I uninstall the Cam scanner app(Android users)?

Prior to uninstalling the app you should first take backup of your important scanned data in the app. 

Steps To take backup of your files from Camscanner:

Click on the link and follow the steps: https://www.camscanner.com/mobile/faq?type=sync_and_backup&id=62

Till now the app was running smoothly but in recent versions of the app the security researchers have found malicious malware which was contained in app's advertising library, this malware is known as Trojan-Dropper.AndroidOS.Necro.n. So what this Trojan malicious code does is downloads more malicious codes into your smartphone and make you sign up into paid services without your knowledge!!

You should delete the app from your phone until the app developers officially announce and upload a stable malware free version on the play store.

• There are many old versions of the app on play store like Cam Scanner HD etc. but you should avoid downloading such versions as it may contain the malicious code.
• The ios users don't need to worry about this as due to apple's strict privacy policy the app won't affect the apple users and it's malware free.

What if i need the stable malware free version ?

If you still need malware free version now, download the apk file provided by camscanner. 
Download the malware free apk file now by clicking this link- https://t.co/79tt0IrMdw

Official Statement from Camscanner:

Steps To take backup of your files from Camscanner:

Click on the link and follow the steps: https://www.camscanner.com/mobile/faq?type=sync_and_backup&id=62

All users should be cautious about the usage of app if you have scanned your sensitive documents please move them to different location than delete the app, your data is your identity.

FaceApp : Good or Bad!!

Everything you need to know!!!!!

Almost everyone who has a smartphone must have heard of FaceApp. Basically it's an App which makes you look old. This App first went viral in 2017 but now in 2019 it came into limelight it has become viral sensation overnight. Almost every person has use this app or must have come across his/her friend or relative's transformed picture. The app has been downloaded millions times in just a couple of days.

Even many celebritries posted their pictures on instagram and twitter and used the hashtag #FaceAppChallenge and this way the app gained emense popularity.

The real question is how this app performs so well and how it gives such accurate results? You'll get to know the answer in the end ;)

What is FaceApp?

FaceApp is a free mobile photo editor app, it makes a person look older/younger with just a click of a button. You can download the app from Google Play Store or Apple App Store. There are some premium features in the app for which you need to pay, but you can look old for free!!

How to use FaceApp?

What you see on the app is just a simple UI consisting of :

Click a selfie or choose a photo from your phone's gallery.

How FaceApp works?

• In layman terms when you click a selfie or choose it from gallery Faceapp upload that photo in it's server and uses the Facial algorithm to process the filters you select.
• The app is being trained with millions of photos of old age people.It uses server side technology to function. In more technical terms it does nueral tranformation in your face ie. smile/not smile, male/female, young/old etc.
• In Machine Learning terms it uses deep learning to give highly accurate results.In one of the article it was mentioned that FaceApp uses "deep generative convolutional neural networks".
• It's hard to predict the exact algorithm/technique as the plus point the makers have is they got millions of face data when the app went viral(back in 2017) and eventually they trained the algorithm with these data to give more precise results about you being old.

It's AI that gives you those wrinkles with a pinch of filter!!

(It's the data on which the algorithm is trained on, which makes it so accurate!! (Answer to the question )

Do I need to worry about my data ?

Unlike other photo editor apps FaceApp doesn't run locally on your phone ie. it uploads your pic of it's server their it uses all those aging/deaging filters and than it gives you your older version.So basically you yourself are allowing the app to use your image.

One of the privacy policy of FaceApp says: (They've removed this 'User contect' privacy policy from their site!!!)

Privacy Policy Source: https://www.faceapp.com/privacy-en.html

How Can I delete my pic?

The app maker say they delete almost all pics (not all) so what they do with other pics??

If a user want's his picture to get deleted. He needs to send a delete request via the app using below steps:

1. Settings
2.  Support 
3. Report a bug 

It's important that we take a note of their privacy policy which says:

"FaceApp may share User Content and your information (including but not limited to, information from cookies, log files, device identifiers, location data, and usage data) with businesses that are legally part of the same group of companies that FaceApp is part of, or that become part of that group (Affiliates)"

Though they have written it "may" share the data, who know's what they do with the users data?? 

Training the algorithm is one of the objective for enhancing app's performance but  nowadays the face data can be used for illegal part as well. There are technologies which takes certain data points from your face and that help to create the faceid dummy and it'll help to gain access where their is facial recognition system.

It depends on you as a user weather you want to get old happily or you want to age using an app filter ;) 

   

Best Data Analytics Documentary you should watch now!!

"Data is the new Oil ? NO: Data is the new soil"

-David Mccandless

Do you know how much data we create every single day?

- It's over 2.5 quintillion bytes of data and it's continuosly growing at an unimaginable speed.

Last week I watched this eye-opener documentary on Netflix - 'THE GREAT HACK'. It's a must watch data analytics documentary on how a company used data to create bias among the people and change their perception using data analytics.Being into the field of data i feel it's very important that everyone should be aware of the consequences of how your data can be misused if you knowingly/unknowingly share your personal information publically.

This documentary shows how a company used more than 5000 data-points(In analytical context any single piece of information is a data point) to hack peoples mind in the US election, so the person will vote for the candidate he didn't wanted to vote and the company achieved this by the buzz word we all have heard of - Data Analytics .Sounds creepy right!!

In a broader sense it's like a psychological warfare where you persuade people by using communication as a means and that too using their own data with their own consent! The documentary shows how the company used the data points of people from the social media datasets and created content by doing Analytics on that data than they used it to persuade them ie. used it for biased politics in the US election.

Let me explain in layman terms how it works!!  

You take the data from a person via social media feed, survey etc.  and  using data analytics you connect all the data points for this person like (likes, dislikes, interests, etc.),than you generate the content which are displayed in the form of ads and you continuosly trigger these ads in that particular person's feed and lastly this changes the perception of the person and he/she is not even aware that they are the victim here. And this is how you persuade the people to achieve the goals via technology such as machine learning, data analytics, data science. 

Things you should remember when it comes to data:

• It's high time we need to think twice before we click on any suspicious link. 

• Never post any of your confidential data on any social media platform. 

• Don't fill any survey form before checking it's authenticity (you'll be giving your data yourself). 

• Stop downloading any app that asks for irrelevant permissions. 

You'll get to know more of these facts when you'll watch the documentary. About the authenticity of how true the events are! All the whistle-blowers are former employees of the accused firm like Data Scientist, Business Analyst etc. 

Will keep you updated with such Data related posts, share this if you find it interesting. Till than bye.

Documentary link - https://www.netflix.com/in/title/80117542