The virtual collaboration is fledging and shifting from the stationary applications more and more into the network. In technical jargon, this trend is referred to as cloud-based collaboration. CloudComputing Insider sheds light on what lies behind it and which concepts and user scenarios make sense.
The cloud is evolving more and more from pure data storage to a collaboration tool that facilitates team and site collaboration. “Offers such as Exchange Online 2010 or SharePoint 2010 support small and medium-sized companies,” says Dirk Backofen, Head of Marketing Business Customers Telekom Deutschland. Countless solutions available on the market now provide customers with practical benefits.
“On the one hand, companies benefit from the growing mobility of their employees, who can access their data anywhere and at any time,” continues Dirk Backofen. Another advantage lies in the cost savings, since the software or infrastructure only needs to be rented and not bought anymore. “Other plus points are greater flexibility, time savings, more homogeneous IT landscapes and complete cost transparency,” concludes the expert.
Legal safe framework for cooperation in the cloud
However, selecting the right technical framework poses some challenges for companies. These already start with the selection of a suitable solution partner. Choosing the wrong provider, you take high risks, especially in terms of security . “Companies should ask themselves, where the data are ultimately stored and processed, in Germany, in the EU or in non-European countries,” confirms Dirk oven from Deutsche Telekom.
Further questions are dedicated to the fine print before the company takes a concrete project target: Are the terms and conditions also in German? Will be settled in Euro? Are the data centers certified? Which data protection and data security measures are applied? Will I get the data back if I cancel the contract? What availabilities are guaranteed during operation? What are the Service Level Agreements ( SLAs )? And last but not least: Will the provider still be in a few years, is it financially strong enough?
The conclusion of Dirk Backofen: “Every company should be aware of its core requirements and closely examine whether the provider can meet them.”
Perceptron: origin and history
The algorithm was invented in 1957 by Frank Rosenblatt , within the Cornell Aeronautical Laboratory. His research was funded by the United States Office of Naval Research and inspired by the cognitive theories of Friedrich Hayek and Donald Hebb. Initially, it was supposed to be a machine rather than a program.
Its first implementation was done as software for the IBM 704, but it was then implemented in a machine created specifically for the occasion called Mark 1 . This machine was designed for image recognition, and consisted of 400 photocells connected to neurons. Synaptic weights were encoded in potentiometers, and weight changes during training were performed by electric motors. This machine is one of the first networks of artificial neurons.
The algorithm was designed to classify visual inputs, categorize subjects into one of two types, and separate groups by one line. Classification is indeed an important part of machine learning and image processing . Machine Learning’s algorithms find and classify patterns in different ways. The perceptron algorithm on its side classifies the patterns and groups them together by finding the linear separation between different objects and patterns received via visual or digital inputs.