How a Machine Learning model progresses from an experiment to an operationalized Web service

Azure Machine Learning Studio provides an interactive canvas that allows you to develop, run, test, and iterate an experiment representing a predictive analysis model. There are a wide variety of modules available that can:

  • Input data into your experiment
  • Manipulate the data
  • Train a model using machine learning algorithms
  • Score the model
  • Evaluate the results
  • Output final values

Once you’re satisfied with your experiment, you can deploy it as a Classic Azure Machine Learning Web service or a New Azure Machine Learning Web service so that users can send it new data and receive back results.

In this article, we give an overview of the mechanics of how your Machine Learning model progresses from a development experiment to an operationalized Web service.

While Azure Machine Learning Studio is designed to help you develop and deploy a predictive analysis model, it’s possible to use Studio to develop an experiment that doesn’t include a predictive analysis model. For example, an experiment might just input data, manipulate it, and then output the results. Just like a predictive analysis experiment, you can deploy this non-predictive experiment as a Web service, but it’s a simpler process because the experiment isn’t training or scoring a machine learning model. While it’s not the typical to use Studio in this way, we’ll include it in the discussion so that we can give a complete explanation of how Studio works.

Developing and deploying a predictive Web service

Here are the stages that a typical solution follows as you develop and deploy it using Machine Learning Studio:

The training experiment

The training experiment is the initial phase of developing your Web service in Machine Learning Studio. The purpose of the training experiment is to give you a place to develop, test, iterate, and eventually train a machine learning model. You can even train multiple models simultaneously as you look for the best solution, but once you’re done experimenting you’ll select a single trained model and eliminate the rest from the experiment. For an example of developing a predictive analysis experiment, see Develop a predictive analytics solution for credit risk assessment in Azure Machine Learning.

The predictive experiment

Once you have a trained model in your training experiment, click Set Up Web Service and select Predictive Web Service in Machine Learning Studio to initiate the process of converting your training experiment to a predictive experiment. The purpose of the predictive experiment is to use your trained model to score new data, with the goal of eventually becoming operationalized as an Azure Web service.

This conversion is done for you through the following steps:

  • Convert the set of modules used for training into a single module and save it as a trained model
  • Eliminate any extraneous modules not related to scoring
  • Add input and output ports that the eventual Web service will use

There may be more changes you want to make to get your predictive experiment ready to deploy as a Web service. For example, if you want the Web service to output only a subset of results, you can add a filtering module before the output port.

In this conversion process, the training experiment is not discarded. When the process is complete, you have two tabs in Studio: one for the training experiment and one for the predictive experiment. This way you can make changes to the training experiment before you deploy your Web service and rebuild the predictive experiment. Or you can save a copy of the training experiment to start another line of experimentation.

The Web service

Once you’re satisfied that your predictive experiment is ready, you can deploy your service as either a Classic Web service or a New Web service based on Azure Resource Manager. To operationalize your model by deploying it as a Classic Machine Learning Web service, click Deploy Web Service and select Deploy Web Service [Classic]. To deploy as New Machine Learning Web service, click Deploy Web Service and select Deploy Web Service [New]. Users can now send data to your model using the Web service REST API and receive back the results. For more information, see How to consume an Azure Machine Learning Web service.

The non-typical case: creating a non-predictive Web service

If your experiment does not train a predictive analysis model, then you don’t need to create both a training experiment and a scoring experiment – there’s just one experiment, and you can deploy it as a Web service. Machine Learning Studio detects whether your experiment contains a predictive model by analyzing the modules you’ve used.

After you’ve iterated on your experiment and are satisfied with it:

  1. Click Set Up Web Service and select Retraining Web Service – input and output nodes are added automatically
  2. Click Run
  3. Click Deploy Web Service and select Deploy Web Service [Classic] or Deploy Web Service [New] depending on the environment to which you want to deploy.

Your Web service is now deployed, and you can access and manage it just like a predictive Web service.

Updating your Web service

Now that you’ve deployed your experiment as a Web service, what if you need to update it?

That depends on what you need to update:

You want to change the input or output, or you want to modify how the Web service manipulates data

If you’re not changing the model, but are just changing how the Web service handles data, you can edit the predictive experiment and then click Deploy Web Service and select Deploy Web Service [Classic] or Deploy Web Service [New] again. The Web service is stopped, the updated predictive experiment is deployed, and the Web service is restarted.

Here’s an example: Suppose your predictive experiment returns the entire row of input data with the predicted result. You may decide that you want the Web service to just return the result. So you can add a Project Columns module in the predictive experiment, right before the output port, to exclude columns other than the result. When you click Deploy Web Service and select Deploy Web Service [Classic] or Deploy Web Service [New] again, the Web service is updated.

You want to retrain the model with new data

If you want to keep your machine learning model, but you would like to retrain it with new data, you have two choices:

  1. Retrain the model while the Web service is running – If you want to retrain your model while the predictive Web service is running, you can do this by making a couple modifications to the training experiment to make it a retraining experiment, then you can deploy it as a retraining web service. For instructions on how to do this, see Retrain Machine Learning models programmatically.
  2. Go back to the original training experiment and use different training data to develop your model – Your predictive experiment is linked to the Web service, but the training experiment is not directly linked in this way. If you modify the original training experiment and click Set Up Web Service, it will create a new predictive experiment which, when deployed, will create a new Web service. It doesn’t just update the original Web service.

    If you need to modify the training experiment, open it and click Save As to make a copy. This will leave intact the original training experiment, predictive experiment, and Web service. You can now create a new Web service with your changes. Once you’ve deployed the new Web service you can then decide whether to stop the previous Web service or keep it running alongside the new one.

You want to train a different model

If you want to make changes to your original predictive experiment, such as selecting a different machine learning algorithm, trying a different training method, etc., then you need to follow the second procedure described above for retraining your model: open the training experiment, click Save As to make a copy, and then start down the new path of developing your model, creating the predictive experiment, and deploying the web service. This will create a new Web service unrelated to the original one – you can decide which one, or both, to keep running.

Linksys WRT32X Gaming Router dengan Dual Band

Sebagai salah satu produsen ternama yang membuat router dibawah kepemilikan Belkin, produk Linksys dimanfaatkan untuk menunjang koneksi data wireless antar jaringan yang berbeda maupun yang sejenis. Dibawah Belkin, Linksys melanjutkan inovasi dalam pasar home router.

Router sendiri mempunyai berbagai jenis tergantung pada fungsinya dalam jaringan perangkat yang ada, tetapi itu tentu bukan merupakan ulasan dalam tulisan di bawah ini.

Jaringan koneksi internet sekarang sudah menjangkau setiap rumah yang ingin terhubung tanpa putus dengan biaya yang terjangkau, membuat wireless router semakin populer untuk dipakai dalam meningkatkan performa trafik data agar bisa menikmati hiburan dan games online tanpa gangguan akibat komunikasi data yang tidak lancar. Ini terutama sangat sensitif bagi para pencinta games online. Segmen ini semakin besar sehingga ceruk pasarnya memang menggiurkan bagi produsen home and small business router.

Dalam menangkap dan memenuhi kebutuhan segmen gaming di Indonesia, Sistech Kharisma, IT solutions Indonesia menghadirkan Linksys WRT32X sebuah gelaran yang diadakan di Upnormal Coffee Roaster pada awal Desember 2017. Rasanya awal 2018 ini menjadi momen yang segar untuk berbagi informasi mengenai spesifikasi dan keunggulan yang ada pada WRT32X.

Peningkatan yang berbeda dengan seri sebelumnya, WRT3200ACM dan WRT1900AC adalah terintegrasinya teknologi Killer Priotization Engine (KPE) dalam firmware, yang memperkuat prioritas trafik data game guna mengurangi puncak ping hingga 77%, meminimalkan isu latensi atau delay, dan akan otomotis mendeteksi PC atau laptop yang sudah terpasang Killer networking card seperti Killer Ethernet (E2500 dan E2400) dan Killer Wireless (AC1535 dan AC1435) serta teknologi gabungan keduanya yakni Killer Double Shot Pro. Hal ini perlu dipastikan untuk memanfaatkan kemampuan yang ada pada Linksys gaming router ini.

Sebagai router dengan dual band (2.4 GHz dan 5 GHz), Linksys WRT32X sudah siap untuk memanfaatkan keuanggulan jaringan 5 GHz yang memiliki spekstrum sinyal yang lebih luas dan tidak memiliki banyak gangguan, dari perangkat elektronik yang banyak bekerja pada frekuensi 2.4 GHz. Kebutuhan bandwith yang besar dalam hal bermain game, streaming film atau transfer file akan terpenuhi dalam skenario ini.

Prioritas data gaming disetel pada panel menu WRT32X dengan mengaktifkan KPE. Menurut Rivet Networks yang menjadi penyedia solusi Killer Networking, seperti dikutip dari CNET.com, berikut adalah urutan prioritas otomatisnya.

Apabila Anda adalah seorang gamer professional atau seseorang yang menggeluti gaming sebagai kompetisi maupun ingin tampil unggul dibanding gamer lain, Linksys WRT32X merupakan pilihan ideal sebagai perangkat pendamping.

Dengan warna hitam tegas, 4 antena pendek dan ukuran yang kecil, 24,5 cm x 19 cm dan tinggi 5 cm, Linksys router ini memberikan sebuah tampilan atraktif di atas rak, kilau LED biru di depan untuk beberapa indikator bisa kedap kedip karena proses internal, namun dapat juga dimatikan melalui menu router.

Microsoft’s sales of Office 365 beat out traditional Office for first time

Revenue from business sales of the Office 365 productivity suite topped traditional corporate Office sales for the first time last quarter — a milestone in Microsoft’s transition from selling packaged software to cloud-computing products.

Seattle Times technology reporter

Six years later, new Office has dethroned the old Office.

Revenue from business sales of the Office 365 productivity suite topped traditional corporate Office sales for the first time last quarter, Microsoft said, a milestone in the Redmond company’s transition from selling packaged software to cloud-computing products.

Microsoft first started selling its iconic collection of Word, PowerPoint, Outlook and other workplace software through online subscriptions under the Office 365 name in 2011. At the time, Google’s Docs and other productivity software were making inroads in a realm then dominated by Microsoft, threatening what had been the company’s most profitable line of business.Microsoft Chief Financial Officer Amy Hood said in a quarterly earnings call Thursday that during the three months ended in June, sales of Office 365 to businesses surpassed sales of the traditional packaged version that entitled the user to the software in perpetuity.

The company didn’t break out the scale of each, but sales of Office products of all sorts to businesses totaled $5.8 billion during the final quarter of Microsoft’s fiscal year. Office 365, analysts say, makes up most of Microsoft’s “commercial cloud,” the line of business-focused subscription products the company has staked its future on. Sales of that group of products stood at $4.75 billion in the most recent quarter, helping to propel Microsoft shares to a record high as investors bet the company is positioned to thrive in the emerging world of web-delivered software.

 Consumers, on the other hand, still spend more on Office licenses than the subscription edition. Those lines of businesses notched about $850 million in sales during the last quarter.Office 365 commercial sales grew 43 percent during the quarter from the previous year. Traditional Office license sales fell 17 percent, a decline Microsoft attributes to companies opting to buy into the cloud version instead. Microsoft has said that it expects to make more revenue per Office customer with the online model.

Office 365 business subscriptions sell for $100 to $420 a year depending on the package, compared with $70 to $100 for the consumer products. Those figures can vary in sales to large businesses, or in packages that include Windows or other Microsoft software.

For those interested in a perpetual license, Office Home 2016 costs $150, and Office Professional 2016 sells for $400.

Cara Office 365 Melindungi Data Penting dalam Perangkat Mobile Anda

Sebagai pengusaha pintar yang sering bekerja secara mobile, perangkat smartphone atau tablet menjadi sangat penting. Dengan perangkat tersebut, Anda melakukan pekerjaan atau bisnis dimanapun dan kapanpun Anda mau. Selain itu, Anda pun bisa menyimpan berbagai data berkaitan dengan pekerjaan atau bisnis Anda tersebut. Salah satu cara melindungi data penting Anda di smartphone atau tablet adalah dengan menggunakan Office 365. Sebenarnya, bagimana cara Office 365 melindungi berbagai data penting Anda?

Menggunakan Berbagai Sistem Keamanan

Untuk melindungi semua data penting Anda, Office 365 hanya bisa digunakan pada perangkat yang legal dan memenuhi standar resmi. Selain itu, layanan Office 365 juga memberikan saran jika data tersebut sebaiknya menggunakan password. Data penting Anda juga akan dilindungi oleh Azure Active Directory Premium. Intinya, data Anda tidak akan mudah dibuka jika bukan Anda yang membukanya.

Memeriksa Data yang Masuk ke Perangkat Anda

Ada kalanya Anda harus mengambil data secara online. Itu artinya akan ada URL yang masuk yang tidak semuanya aman. Untuk memberikan perlindungan, Office 365 menggunakan 2 fitur yaitu URL Detonation dan Dynamic Delivery. Dengan dua fitur pengamanan ini, Office 365 akan melakukan pemeriksaan terhadap link yang masuk ke perangkat smartphone atau tablet Anda seperti data dalam bentuk Office, PDF, dan lainnya. Fitur pengaman ini akan memeriksa apakah file tersebut berpotensi membawa virus atau tidak. Jika memang berpotensi berbahaya bagi smartphone atau tablet Anda, fitur tersebut akan langsung menghalangi akses link tersebut. Dengan fitur Dynamic Delivery, Anda tetap bisa membaca file atau data saat Office sedang melakukan pemeriksaan pada file yang terlampir. Untuk menggunakannya juga sangat mudah. Cukup aktifkan kedua fitur tersebut maka keduanya akan bekerja secara otomatis.

Melindungi Data Penting Anda

Ada kalanya Anda harus melampirkan identitas atau informasi penting namun sifatnya pribadi. Misalnya, Anda harus mencantumkan nomor kartu kredit atau nomor-nomor lainnya. Salah satu fitur Office 365 akan melindungi data penting tersebut sehingga tidak sampai jatuh ke tangan yang tidak bertanggung jawab. Jika Anda menerima data atau file dalam bentuk Exchange Online, SharePoint Online, atau One Drive, maka fitur tersebut akan langsung melakukan pemeriksaan secara otomatis. Anda juga akan mendapatkan peringata jika data tersebut berpotensi berbahaya bagi perangkat Anda.

Adanya Fitur Rights Management

Satu lagi fitur pelindung yang ada dalam Office 365 adalah Office 365 Rights Management. Fitur ini berfungsi untuk melindungi semua data-data atau email yang ada di perangkat smartphone atau tablet Anda. Fitur ini memungkinkan Anda untuk melihat dan mengatur data yang ada tersebut. Fitur ini melindungi data Anda dengan cara mengelompokkan data yang ada, memberikan label pada data tersebut kemudian memberikan perlindungan. Untuk lebih memundahkan Anda dalam memilah data-data tersebut, Office 365 memperbolehkan Anda untuk memberikan klasifikasi secara manual.

 

Microsoft Azure

Empat hal yang harus dimanfaatkan oleh bisnis kecil dan menengah dari transformasi digital

Mengenal lebih dekat si pintar Microsoft Office 365

Solusi Azure Backup Untuk UKM