Backpropagation Neural Network(formula)
Count triangles on GraphX
As the official guideline of GraphX, Official GraphX triangle counting, the function that counting triangles is limited for the edges should have to be in canonical orientation (srcId < dstId). However, triangles is important in a graph especial social network. In this case, i implemented a better way to calculating triangles on GraphX.
Pregel: Spark GraphX
Owing to large scale graph is the primary concern when considering a practical computing problem. Furthermore, the scale of graphs is too large which almost contains billions of vertices and trillions of edges to run in a single machine. Based on this circumstance, utilizing parallel and distributed computing are regarded as a better solution. GraphX, a commonly employed distributed data-flow system which focusing on graph processing building on Apache Spark, is a better way to address massive graph problem (Gonzalez, Xin, Dave, Crankshaw, Franklin and Stoica, 2014). Then how to build a parallel program using GraphX?
Research on Automatic License Plate Recognition based on Character Segmentation
Introduction
With the rapid development of Computer Science, automatic license plate recognition, i.e. an applied system which can operate and distinguish license plates by itself, are regarded as a symbol of the trend that humans tend to utilize computer programs to address problems more than ever. It is widely employed for helping other systems, e.g. calculating costs of parking or tracing vehicles (Mohindra and Shukla, 2013).
Feedback of the introduction of R
This is an article talking about feedback of the introduction of R
Content of R
As the course name Coursera recommended, R is an advanced product based on S language. S language was developed by John Chambers and others at Bell labs. The basic philosophy of S have two respects. The initial one is that human can utilizing the language to check and analysis data in an interactive environment without knowing very detailed aspects, or having to know how to make it. Furthermore, the other one is that it is an opportunity for people who wants to develop their own tools and the system by outgrowing this language. Owing to R language inherit S language, these principle of S suitable for R as well.
Literature review of Comparing group recommendation algorithms
Introduction
With the rapid development of computer science, adopting mathematics to address problems gradually became a convenient method over the past several decades. In such environment, recommendation algorithms make contribution to expand the approach of selling goods. Along with the appearance of Big Data and Cloud Computing, experts have begun to pay more attention to the utility of mathematic algorithms. From the late nineties, the year of the first scientific publications introducing group recommendation systems, many researchers believe that group recommendation can be generated by collecting individual recommendations into a recommendation group or by collecting personal preference into a preference group(Pessemier, Dooms and Martens, 2013).
Since customers usually do not know what they want especially when they are surfing the Internet, which is the most severe challenge for a recommendation system, it is really hard to demand an algorithm to satisfy every request. The initial group recommendation system named Content-based recommendation algorithm was built in 1994 and gave out an recommendation like a personal prediction (Pagare and Patil, 2013). A year later, Collaborative filtering recommendation algorithm, a more advanced algorithm, was regarded as a more effective solution when the previous algorithm met problems. The last candidate is called the Hybrid recommendation algorithm which is the latest algorithm. In terms of the usefulness and compatibility of recommendation algorithms, this literature review will focus on comparing several algorithms to reveal not only the basic theory of recommendation algorithms and machine learning but also to distinguish which algorithm might be the best solution in a specific area.
OY.Troy 