Python For Data Science

New Articles on Python for data science.

Approaches to Augmented Reality

Augmented reality is a technology that enriches the real world with digital information such as GPS data, 3D models, graphics, audio or any other computer-generated sensory inputs by superimposing them on real-world objects seen through a camera on your smartphone, PC, or connected glasses.

Augmented reality is a bridge between real and virtual realms which helps us to understand the world around us.

1) Marker Based AR

It uses black and white markers which serve as activators for additional information when a camera is pointed at an object. These markers are images, such as QR or 2D codes, which contain information and prompt some action when they are sensed by software.

Marker-based augmented reality has been widely used in advertising, such as paper catalogs, magazines, and posters. 

2) Sensor Based AR

It doesn’t require black and white markers. Instead, sensor-based augmented reality uses the Internet, cameras, and a whole range of optical and other sensors such as GPS, compasses, accelerometer and gyroscopes that are built into devices. With sensor-based AR, you can point your phone in any direction and see the world augmented with image and media overlays. for example: Snap chat.

Augmented Reality Tracking Technologies

1) Image tracking

In this, you need a photo or marker which contains a code and can be compared to a QR Code. Scanning the code you can admire the visualization of your content in 3D. This is the easiest way to create a tracking pattern for your app.

2) GPS/orientation tracking 

This might be useful for companies, shops or tourist destinations. The user can display any location-related information, for example, shops nearby.

3) Object Tracking

It allows using real objects as targets. You need to deposit the 3D data of your target. The app is able to scan not only planar images but also complex 3D objects independently of their size and geometry.

4) Markerless Tracking

In this, you scan the object and sensors within your device measure the object’s direction. Without lodging 3D data you can visualize your content due to the sensors that measure the direction of your tracker.

How YouTube's Recommendation Algorithm Works

YouTube recommendations are driven by Google Brain, which was recently open sourced as TensorFlow.

link:       https://www.tensorflow.org/

By using TensorFlow one can experiment with different deep neural network architectures using distributed training.The system consists of two neural networks.

The first one, candidate generation, takes as input user’s watch history and using collaborative filtering selects videos in the range of hundreds. An important distinction between development and final deployment to production is that during development Google uses offline metrics for the performance of algorithms but the final decision comes from live A/B testing between the best performing algorithms.

Candidate generation uses the implicit feedback of video watched by users to train the model. Explicit feedback such as a thumbs up or a thumbs down to a video are in general rare compared to implicit and this is an even bigger issue with long-tail videos that are not popular. To accelerate training of the model for newly uploaded videos, the age of each training example is fed in as a feature. Another key aspect for discovering and surfacing new content is to use all YouTube videos watched, even on partner sites, for the training of the algorithm. This way collaborative filtering can pick up viral videos right away. Finally, by adding more features and depth like searches and age of video other than the actual watches, YouTube was able to improve offline holdout precision results.

The second neural network is used for Ranking the few hundreds of videos in order. This is much simpler as a problem to candidate generation as the number of videos is smaller and more information is available for each video and its relationship with the user. This system uses logistic regression to score each video and then A/B testing is continuously used for further improvement. The metric used here is expected watch time, as expected click can promote clickbait. To train it on watch time rather than click through rate, the system uses a weighted variation of logistic regression with watch time as the weight for positive interactions and a unit weight for negative ones. This works out partly because the fraction of positive impressions is small compared to the total.

YouTube’s recommendation system is one of the most sophisticated and heavily used recommendation systems in the industry.

Difference between E, H, H+ and 3G that appear when we turn our mobile data on?

• E is edge (also called Enchanced GPRS) which is a enhancement over gsm/gprs (2G) technology. Data rates of around 400kbps. It basically uses gsm/gprs infrastructure.
• 3G is the next major technology which uses WCDMA radio signalling.  The data rates here are comparable to edge but It offers better quality, better latency, lower delay, so making it more suitable for audio, video and streaming services.
• H is HSPA which is high speed packet access.Its a enchancemnet over 3G and is mix of HSDPA (downlink) and HSUPA (uplink). Major changes were done on radio side which increased the data rates upto 3 and 7Mbps (Theoretically its much more)
• H+ is Evolved HSPA. HSPA was further improved into HSPA+ where technologies like MIMO antennas were introduced which boosted the data theoritically around 160Mbps downlink and around 15Mbps uplink (practically data rate is as we see this advertised as 21Mbps)

If you subscribe to a 3G plan you have then subscribed to all the above. Your mobile first looks out for cells/tower which are H+ capable, if it fails it tries the lower ones. When 3G is not at all available mobile falls back to Edge.

Also if you observe when you are in Edge, your calls and data will not work together. This is because voice and data calls use the same infrastructure in 2G whereas in 3G (or H or H+) you will be able to able to browse on your phone while you are on call. This because data and voice are separated and can work parallely in a 3G network.

Khoobsurat

Ke bahut khoobsurat ho tum.
Is Dil ki dharkan ho tum...
Lekin mere bagaer adhuri ho tum

Kisi ki dost ho tum.
Toh kisi ki aashiqui ho tum.
Lekin mere liye....
meri zindagi ho tum.

Best Books on Psychology

1)                                                                                                                         4)

       



2)                                                                                                            5)
           
                                                                                                                     


3)