Comparison of 5 Analytical Tools
Short Description
Analytic tools such as IBM SPSS, Tableu...
Description
IBM SPSS Modeler By integrating predictive analytics with decision management, scoring and optimization in organization's processes and operational systems, SPSS Modeler helps users and systems make the right decision every time. Important features of SPSS Modeler: Analytical decision management Automate and optimize transactional decisions by combining predictive analytics, rules and scoring to deliver recommended actions in real time. Decision management capabilities enable the integration of predictive analytics and business rules into an organization’s processes to optimize and automate high-volume decisions at the point of impact.
Automated modelling Use a variety of modelling approaches in a single run and then compare the results of the different modelling methods. Select which models to use in deployment, without having to run them all individually and then compare performance. Choose from three automated modelling methods: Auto Classifier, Auto Numeric and Auto Cluster.
Text analytics Go beyond the analysis of structured numerical data and include information from unstructured text data, such as web activity, blog
content, customer feedback, emails and social media comments. Capture key concepts, themes, sentiments and trends and ultimately improve the accuracy of your predictive models.
Social network analysis Social network analysis examines the relationships between social entities and the implications of these relationships on an individual’s behaviour. It is particularly useful for those in telecommunications and other industries concerned about attrition (or churn). By identifying groups, group leaders and whether others will be affected based on influence, predictive models can be built on an individual and enhanced with their group and social behaviour data.
The modelling algorithms included in SPSS Modeler are: Anomaly Detection. Detect unusual records with a cluster-based algorithm. Apriori. Identify the frequent individual items in your transactional databases and extend them to larger item sets. Bayesian Networks. Estimate conditional dependencies with graphical probabilistic models that combine the principles of graph theory, probability theory, computer science and statistics. C&RT, C5.0, CHAID and QUEST. Generate decision trees, including interactive trees.
CARMA. Mine for association rules with support for multiple consequents and continuous feedback for deterministic and accurate results. Cox regression. Calculate likely time to an event. Decision List. Build interactive rules. Factor/PCA, Feature Selection. Reduce data. Generalized Spatial Association Rule: Find patterns/association rules where location matters. K-Means, Kohonen, Two Step, Discriminant, Support Vector Machine (SVM). Cluster and segment data. KNN. Model and score nearest neighbour. Logistic Regression. Generate binary outcomes. Neural Networks. Take advantage of multilayer perceptron’s with backpropagation learning and radial basis function networks. Regression, Linear, GenLin (GLM), Generalized Linear Mixed Models (GLMM). Model linear equations. Self-learning response model (SLRM). Take advantage of a Bayesian model with incremental learning. Sequence. Conduct order-sensitive analysis with sequential association algorithm. Spatial-Temporal Prediction (STP). Predict how a place will change over time. Support Vector Machine. Apply non-linear functions based on computational learning theory for efficient learning on wide datasets. Time-series. Generate and automatically select time-series forecasting models using techniques such as temporal causal modelling, which discovers causal relationships among large numbers of series. Two-step clustering: Identify data points by similarity and group them into clusters.
TABLEAU In today’s competitive world organizations are expecting a solution that can satisfy their information needs in just one click; solve business queries in seconds; with option to customize as per organizational needs. What Information Technology offers is a range of Business Intelligence Tools that assist organizations in taking business decisions more effectively and efficiently.
Out of 100s of Business Intelligence Tools available in the market, I experienced tableau and believe me it’s just awesome. This BI Tool comes with many rich features. You simply have to drag and drop things and you will get what you actually want, that too in just fraction of seconds. Tableau offers four business intelligence products viz.
Tableau Desktop
Tableau Server
Tableau Digital
Tableau Public
Key differentiators
Key Features of Tableau Tool for everyone Most often, it has come to notice that, there exists a gap between the report developer and business decision maker, i.e. the one who consumes the report. Tableau bridges this gap to a great extent by offering report building capability which a lay man can use to develop a Business Intelligence Solution. This helps in minimizing the dependency on IT resources. The primary and one of the best features of Tableau is you need not be a technical expert. Business users like decision makers, department heads, sales and marketing heads, sales representatives, senior management can easily and effectively use Tableau to develop and generate reports to satisfy their information need while taking business decisions. Fast Tableau is one of the fastest Business Intelligence Tool available in the market. Tableau can create reports in minutes which might take hours in Microsoft Excel. One can connect to data source and in few minute’s time one can have attractive graphical visualizations to choose from. Maps
Tableau’s in-built geocoding feature makes Map a core part of Tableau. Whenever users have data related to common areas like countries, states or post codes, they need not enter latitude or longitude data to locate that area on the map. Along with strong data visualization, tableau gives us flexibility to tell stories with maps and to use maps to drill down into related information.
Interactive Data Visualization
Tableau is one of the interactive Business Intelligence Reporting Tools available in the market. Tableau understands the type of data connected and suggests suitable visualization. Features like “Show Me” recommends the suitable data visualization based on the selected data. Once it is connected to a database, Tableau automatically segments data into dimensions and measures. Tableau offers vast number of data visualizations by reading the data which helps in discovering significant business findings. Tableau also offers drilling through the data, hence making it possible to reach to the lowest and most detailed information possible. Flexibility Tableau provides flexibility to choose how you work with data. One can directly connect to database or can use in memory technology. In memory can be used when database is slow or one needs to work offline or when live database connection is not possible. Tableau has an ability to connect directly with almost all types of databases available in the industry. This significantly helps in reducing time and efforts for aggregating the data and importing into local data warehouse. Having direct connectivity with various data sources, more time can be spent in analysis. Tableau also understands and extends all the native functions available in the connected database. Economical Tableau is one of the most economical Business Intelligence Tools available in the market. Organizations can also save money as there is less dependency on additional IT resources.
Sharing of information Tableau Server offers offline and online publishing of the data visualization which makes the information exchange fast, simple and effective. Tableau Server helps to share information along with underlying summary data via emails along with various data exports formats. Maintains Security Tableau respects security protocols. Every user has a unique id and password that restricts the access to business information. Gartner’s Report
Magic Quadrant for Business Intelligence Platforms As per Gartner’s Magic Quadrant for Business Intelligence Platforms report 2012, tableau is placed under “Challenger” based on Gartner’s assessment of Tableau’s “ability to execute” and “completeness of vision”. According to Gartner’s report, Tableau has gained overwhelmingly positive customer survey feedback across the board in all measures in the survey, including ease of use, functionality, product quality, product performance, support, customer relationship, success, achievement of business benefits and view of the vendor's future
IBM COGNOS ANALYTICS REVIEW Product Overview
IBM Cognos Analytics software is an online-based business intelligence platform that offers a complete array of BI software to address company goals. Considered one of the leaders in the field of business intelligence, Cognos Analytics is comprised of over thirty different products. Cognos’ abilities include dashboarding, reporting, analytics, scorecarding, and, notably, data integration. Among its greatest strengths is its attention to customer feedback in the design of its products, consistently earning it high scores as a manufacturer that understands the needs of its clientele very well. Its advanced analytical decision making software has especially been noted, allowing for organizations to automate and strengthen sound business decisions. Cognos Analytics also contains unique products that completely separate it from the broader field of business intelligence by being able to simultaneously cater to both the largest corporate giants on the one hand, and smaller to midsize companies on the other, with the same system. Cognos Express is designed specifically for departments within larger companies, and smaller and midsized companies, that are unable to afford exorbitant implementation fees. Express is limited to 100 users and has all the same reporting, querying, dashboarding, and scorecarding capabilities of the full-scale software. This allows businesses that before now have been largely unable to afford the technology to use the same analytical processes as its much larger competitors. In October 2015, Cognos has been redesigned to utilize the same selfservice functionality as its new Watson Analytics.This greatly benefits businesses as they won’t need to rely on their IT staff to compile BI reports. They can access dashboards and run reports from any device. Other new features include:
Intent-driven modeling: Non-technical users are able to pull their data together without having to understand the technical stuff that goes with it. Cognos Analytics can interpret the intent of the attributes that’s typed, search the data sources that best match the attributes and suggest some data modeling scenarios.
Smarter search functionality: The smarter search functionality helps users find the right information quickly. For example, smart search can guide the user to add data to charts or provide visualization tools relevant to that search.
Single platform for all types of business reporting: All of Cognos Analytics’ reports and dashboards are in one environment and accessible from any device. Employees that work outside of the office are able to access the same data and tools as those in the office, which makes for easier collaboration.
Features
Cognos Analytics is usable with multidimensional and relational data sources from various companies, such as Oracle, SAP, Microsoft, and more. Featured programs include: Cognos TM1, Cognos Insight, Cognos Express, Cognos Enterprise, Cognos Disclosure, Management. The Cognos Express debuted its latest version in June 2015 with these new features/enhancements (those that own this program will now have access to the latest versions/releases of Cognos TM1 and Cognos BI):
Simpler user interface and deployment
Ability to install Cognos TM1 and Cognos BI on separate machines
Ability to work with Microsoft Excel
SSL Support
Linux support for Cognos TM1
Analysis Features
Ad Hoc Analysis
Online Analytical Processing (OLAP)
Predictive Analysis
User Friendly Interface
Reporting Features
Ad Hoc Reporting
Automatic Scheduled Reporting
Customizable Dashboard
Customizable Features
Graphic Benchmark Tools
Target Market IBM services over 23,000 companies from a broad cross-section of industries including aerospace, defense, banking, education, healthcare and many more. Below are some of their clients:
Nike
GKN Land System
Spain's Ministry of Defense
British Airways
Chemring
Quinte Health Care
Troy Corporation
Michigan State University
Lufthansa Cargo
Jabil
Shortcomings Critics of the product cite its difficulty to use, especially for those new to advanced software. Of particular note are the error messages that continually pop up, and have been noted to be very difficult to cipher and even more difficult to resolve. Data reports also take almost twice as long to compile with Cognos Analytics as compared to most competitors. A few users have mentioned that there’s too many components that needed to be installed.
RapidMiner Product Information: RapidMiner is one of the most widely used analytics platforms in the world, with over 250,000 users. Organizations of all sizes use RapidMiner, and its range of application is very broad. The fact that many predictive models can be built without resorting to program code is one reason for its popularity, the other being very reasonable pricing. This is a sophisticated offering with over 1500 drag-and-drop operators. Novice users can quickly get up to speed with RapidMiner’s ‘Wisdom of Crowds’ online repository of best practices, and this is quite unique among analytics platforms. Big data is well accommodated through its Radoop platform – insulating users from the complexities and volatility of big data technologies. The importance of this facility cannot be over-emphasized, as organizations struggle to keep up with rapid developments in big data technologies. Organizations of all sizes looking for a cost effective, powerful analytics platform, will find that RapidMiner is a speedy, scalable environment in which to develop and deploy predictive models.
Business Application RapidMiner is used in every conceivable industry, from cement manufacturing through to electronic payment companies. This range of applications demonstrates very well the versatility of the platform, and
both medium and large businesses benefit from its low cost of entry and sophisticated capabilities. One of the most high profile users of RapidMiner is PayPal. It needed to get an inside track on customer churn. The text analytics capabilities of RapidMiner were used to classify customers as ‘top promoters’ and ‘top detractors’ by analyzing feedback. This in turn enabled product managers to take action when negative sentiment was detected. A good example of the result of this analytics was the identification of problems associated with passwords. Changes were made and a corresponding drop in negative comment was the outcome. Body Biolytics provides activity recognition software to the sport, fitness and health industries. Wearable devices generate huge amounts of data, and Body Biolytics uses this to monitor sports and fitness activities. This results in the creation of performance baselines, the tracking of performance improvements, and when a user is over-exerting. The highly bespoke nature of these applications illustrates quite well the diverse uses of RapidMiner, and its flexibility. Technology RapidMiner is a graphical, drag-and-drop analytics platform that is used for creating predictive models. RapidMiner Studio sits on the desktop and allows a data scientist or analyst to build predictive models in a local environment – these can be deployed into production via a number of mechanisms including PMML. RapidMiner Server supports project based work and supports high levels of collaboration. Regardless of variant RapidMiner supports a full development cycle from data wrangling, exploration, model development and testing, and deployment –with no recourse to program code. New and seasoned users of RapidMiner are aided by the online Wisdom of Crowds best practice repository. RapidMiner Radoop allows users to employ big data (or any other data) without having to dive into the complex, changing technologies associated with the Hadoop ecosystem. This means models can be developed in RapidMiner and deployed in the big data environment, utilizing big data technologies (Mahout and MLib for example). RapidMiner Streams utilizes the Apache Storm stream processing capability. Models are developed in RapidMiner and deployed into Storm using pre-built connectors and the application of any of the 1500+ RapidMiner operators. Applications include streaming data from production machines, wearable and IoT devices and financial data streams. RapidMiner Cloud supports the distribution of compute intensive tasks to cloud based facilities. Operators are available to connect with a variety of
social and application data sources, including Twitter, Salesforce, Dropbox and over 300 other cloud data sources.
Productivity is a major push for RapidMiner, and to this end it supplies Accelerators and Wisdom of Crowd. The Accelerators are effectively solution templates (customer churn for example) that give sixty to eighty per cent fit. Wisdom of Crowd is a large repository of analytic best practices from the 250K users. RapidMiner uses a machine learning repository to recommend how to build an analytic process. By guiding the analytic development, both novices and experts gain knowledge on how to make their models perform better.
Teradata Aster Discovery Teradata Aster Discovery products provide tools and technologies that enable users to efficiently perform data discovery tasks using advanced big data analytics on all types of data for the entire data discovery process. These tools and technologies include SQL, MapReduce, graph and data stores for all data types that include structured, semi-structured and unstructured data. Combining several technologies into one platform provides for layers of analytics, thus enabling the discovery of insights in data that would be more difficult to find using individual or disparate tools.
Aster's massively parallel processing (MPP) data warehouse architecture supports embedding applications within the database engine, enabling users to perform fast analysis of massive data sets.
Components of the Teradata Aster Discovery products Aster Discovery Platform provides the framework upon which Aster tools and processing engines are integrated to help users perform advanced data analytic exploration, resulting in improved data insights. Aster Database is the primary data store of the Aster product. It's a highperformance, scalable MPP database that embeds MapReduce analytic processing with data stores for big data analytics on multistructured data sources and types. Aster Discovery Portfolio provides prebuilt analytic functions that can be used for various big data applications, such as graph analysis, path and pattern analysis, predictive statistical analysis, text analysis and SQL
analysis in the Aster Discovery Platform. More than 80 analytics functions are provided that address use cases in a broad range of industries, including fraud and risk analysis, churn, manufacturing optimization, golden path analysis, marketing attribution, sentiment extraction, influencer analysis and personalized recommendations. Aster's Discovery Portfolio is part of Teradata Aster Discovery Platform, which also includes Teradata's Aster Database. Aster SQL-GR is a graph processing engine for performing graph analytics on big data sets in the Aster Database. Graph discovery enables users to analyze complex graph network structures. Connection Analytics is powered by the Aster SQL-GR engine and an expanding set of built-in graph analytics functions. This tool helps discover relationships and connections within a network of products, people or processes. Connection analytics can be applied to business areas such as security, marketing and human resources, and can be used to find top influencer nodes in a graph. Aster SNAP Framework, which was introduced in Release 6.1, enables the different analytical engines and storage engines to easily integrate with the Aster Discovery Platform. SNAP also includes a query optimizer for processing across multiple types of workloads, including Graph, R, MapReduce and SQL. The Aster SQL-MapReduce framework enables users to perform highperformance analytics of MapReduce using standard SQL or R against the Aster Database. Aster R combines the power of Aster's SNAP Framework with the capabilities of the open source R engine. This approach allows R programmers to take advantage of the prebuilt functions provided by Aster's Portfolio product in addition to the capabilities of R libraries. Programmers using R in Aster's framework can achieve massive scalability by leveraging Aster's Database and MPP architecture. Aster Database runs on Red Hat Enterprise Linux or a Terada distribution of SUSE Linux Enterprise Server. It has been certified to run on several commodity platforms that include specific hardware from Dell and HP. Hardware and software configuration requirements can be found in the Aster Database Server Platform Matrix. Teradata also offers Aster Database Cloud, which provides users with the flexibility and agility of cloud computing for big data analytics while leveraging Aster's massively parallel analytics engine. A version of Teradata offers Aster Express, a virtual appliance that runs in VMware Player on a PC. Aster Express is a fully functional cluster that can be used for evaluation, testing or development purposes.
Teradata Aster 6.1 also includes improvements in connectors as well as enhanced integration with other Teradata products such as Teradata Viewpoint and support for new Teradata Studio features.
View more...
Comments