Un/Supervised Learning: From a theoretical point of view, supervised and unsupervised learning differ only in the causal structure of the model. In supervised learning, the model defines the effect one set of observations, called inputs, has on another set of observations, called outputs. In other words, the inputs are assumed to be at the beginning and outputs at the end of the causal chain. The models can include mediating variables between the inputs and outputs.

In unsupervised learning, all the observations are assumed to be caused by latent variables, that is, the observations are assumed to be at the end of the causal chain. In practice, models for supervised learning often leave the probability for inputs undefined. This model is not needed as long as the inputs are available, but if some of the input values are missing, it is not possible to infer anything about the outputs. If the inputs are also modelled, then missing inputs cause no problem since they can be considered latent variables as in unsupervised learning.

Clustering: Cluster analysis is, essentially, the process of taking data and grouping it according to similarities. It might also be called pattern recognition. In the area of supervised classification, we start with pre-determined categories and then label the data accordingly. Cluster analysis is different in that we allow the process itself to identify reasonable categories.

Partitional clustering : K-means is one of the simplest unsupervised learning algorithms that solve the well known clustering problem. The procedure follows a simple and easy way to classify a given data set through a certain number of clusters (assume k clusters) fixed a priori. The main idea is to define k centroids, one for each cluster. These centroids should be placed in a cunning way because of different location causes different result. So, the better choice is to place them as much as possible far away from each other.

Hierarchical clustering: In hierarchical clustering the data are not partitioned into a particular cluster in a single step. Instead, a series of partitions takes place, which may run from a single cluster containing all objects to n clusters each containing a single object.  Hierarchical Clustering is subdivided into agglomerative methods, which proceed by series of fusions of the n objects into groups, and divisive methods, which separate n objects successively into finer groupings.

Measurement of clustering:

• Single link: in this method the distance between two clusters is determined by the distance of the two closest objects (nearest neighbours) in the different clusters. This rule will, in a sense, string objects together to form clusters, and the resulting clusters tend to represent long “chains.”
• Complete link: In this method, the distances between clusters are determined by the greatest distance between any two objects in the different clusters (i.e., by the “furthest neighbours”). This method usually performs quite well in cases when the objects actually form naturally distinct “clumps.” If the clusters tend to be somehow elongated or of a “chain” type nature, then this method is inappropriate.
• Average link: This method is identical to the unweighted pair-group average method, except that in the computations, the size of the respective clusters (i.e., the number of objects contained in them) is used as a weight. Thus, this method (rather than the previous method) should be used when the cluster sizes are suspected to be greatly uneven.
• Centroids: The centroid of a cluster is the average point in the multidimensional space defined by the dimensions. In a sense, it is the centre of gravity for the respective cluster. In this method, the distance between two clusters is determined as the difference between centroids.

Week 12 = Revision week.

Tutorial – Personalized Medicine: Management of a patient’s medical condition on the basis of the use of diagnostic tools employing molecular analysis of the patient’s genome to achieve the optimal medical outcomes for that individual. Personalized medicine is an extension of traditional approaches to understanding and treating illness. Since the beginning of the study of medicine, physicians have employed evidence found through observation to make a diagnosis or to prescribe treatment. In the past, this was presumably tailored to each individual, but personalized medicine makes treatment more specific.

Bioinformatics: The use of computers in solving information problems in the life sciences. It mainly involves the creation of extensive electronic databases on genomes, protein sequences etc. Also involves techniques such as three-dimensional modelling of biomolecules and biological systems. The primary goal of bioinformatics is to increase our understanding of biological processes. What sets it apart from other approaches, however, is its focus on developing and applying computationally intensive techniques (e.g. pattern recognition, data mining, machine) to achieve this goal.

Reference:

[1]: http://www.cis.hut.fi/harri/thesis/valpola_thesis/node34.html

[2]: http://csurs1.csr.uky.edu/~pthacker/cluster.html

[3]: http://www.resample.com/xlminer/help/HClst/HClst_intro.htm

[4]: http://home.dei.polimi.it/matteucc/Clustering/tutorial_html/kmeans.html

[5]: http://www.personalizedmedicine.com/

There are many communication methods such as face to face letter or telephone, as well as new and innovating ways to communicate in healthcare such as video conferencing EDI and some structured LANS and WANS. These methods are all very possible but can be prohibited by certain problems for instance costs vs. benefits and timely set-up cost or smaller problems like ‘will the user actually be able to run and operate certain computerized features of communication. Certain requirements must be maintained to have communication software and hardware in place, addressing and software routing as well as maintenance and terminal connections and security must be implemented correctly. In the lecture we also went through the different tiers of system architecture, Presentation tier (GUI), application tier, data/database tier. These tiers as wells as authentication and security we some of the details that followed in creating communications and networks in health care so that information can be encrypted and verified while remaining private and maintaining confidentiality. We also looked into the International Organization for Standardization (ISO) whom began to develop its OSI framework architecture. “It is an abstract description for layered communications and computer network protocol design. In its most basic form, it divides network architecture into seven layers which, from top to bottom, are the Application, Presentation, Session, Transport, Network, Data-Link, and Physical Layers. A layer that provides error-free communications across a network provides the path needed by applications above it, while it calls the next lower layer to send and receive packets that make up the contents of the path”. [Ref 1: http://en.wikipedia.org/wiki/OSI_model]

In week 8 we studies telemedicine, telecare and patient monitoring systems.

• Telemedicine is thought of as long-distance clinical health care, including practitioner-to-patient meetings (formally referred to as telemedicine encounters), practitioner-to-practitioner discussions (formally referred to as telemedicine consults) and exchange of clinical information via technology. [www.sewtn.com]
• Telecare is the term given to offering remote care of elderly and vulnerable people, providing the care and reassurance needed to allow them to remain living in their own homes [www.Wikipedia.com/telecare]
• Patient monitoring is the continuous observation of patients life support of physiological function for the purpose of guiding management decisions, including when to make  therapeutic interventions, and assessment of those  interventions” [Hudson, 1985, p. 630].

Tutorial:

Biometrics are automated methods of recognizing a person based on acquired physiological or behavioural characteristics.  Biometric technologies are up-and-coming in the health care setting, from access control for doctors to patient identification. Healthcare professionals have stated the significance in using fingerprint recognition as a way of authenticating an persons access to the information systems but must be aware of cleaning regimens of contact devices to minimize the potential for bacterial infections to spread.

What is DICOM – Digital Imaging and Communications in Medicine.

DICOM differs from other data formats in that it groups information into data sets. That means that a file of a chest X-Ray image, for example, actually contains the patient ID within the file, so that the image can never be separated from this information by mistake.

l       Under MATLAB image processing toolbox:

• dicomread(filename);
• dicominfo(filename); get metadata from DICOM file
• dicomwrite(X, filename); write the image X to a DICOM file
• dicomanon – remove confidential medical information
• dicomdict – set dicom directory
• dicomlookup – look up attribute into the dicom directory
• dicomuid – cerate a UID
• 

Using the three-tier architecture we create a basic healthcare webpage to display the information stored in the database that we created in Week6 practical.

Tutorial: In the seminar we covered the problems with NHS and some of the proposed NPfIT solutions. Some of which being the improved patient record sharing, diagnosis retrieval based on patient history and how the MHS email and directory service will enable secure transmission of patient information. Health space – online personal health organiser will allow patients to record their own blood type allergies and other unique details for clinicians. We also looked through GP support systems and NCRS (national care record system) which will hopefully allow for the implementing of an EPR system in the NHS. They claim to provide local integration of clinical systems and revolve around the central data spine and will provide a high level record of clinical data and healthSpace.

Practical: in the practical we were too create a database using ms access more or less report and form retrieval by typing all the details given to using in the practical sheet in tables in the correct format and make queries based on each table to gain proper and required feedback to forms and rpts. Patient table was created first and amended with a number of input masks and a few other little details to story errors and duplications and from the a query was based on the table and a form which designed. I add my own details on the db like a search query action button to allow the user to enter in the surname of patient to gain access to that specific person details as well as gain a report based on that. As always with access relationships were made and ad hoc queries to as user questions. I created these as well as amending and delete query. I made my way through the different questions for creating queries. The search query was pretty simple in SQL

SELECT firstname, lastname, patientnum, address, age, doctornum, phone

FROM tbl_patient

WHERE (“lastname” = [enter patient surname]);

SELECT doctornum, age, firstname, lastname
FROM tbl_patient
WHERE (doctornum=[”2”]);

SELECT doctornum, gender, firstname, lastname
FROM tbl_patient
WHERE (gender=[”female”]);

GROUPBY

• Week 3: I this lecture we watched a few different videos on how x-ray technology and machinery is used and how it works as well as other scanning technologies in health such as CAT, CT, MRI, Ultrasound, ECG, EEG, BCI, EMG, EOG. We gained information on how each of these technologies work and use specialist signals to make diagnosis as well as the applications involved.
• Week 4: in this lecture we went through information processing of medical data in different stages e.g. observation, diagnosis, and therapy. We researched different types of medical data as in visual images, analogue, narrative as well as PACs and 3d medical imaging and simulations and finally e-prescribing.

is a medical practitioner’s patient-specific clinical and economic information, for consenting patients, to prescribe the most medically appropriate and cost effective prescription at the point of care.

Tutorial

• Week 3: the tutorial in was a chance for us to go over the practical from week two were we had a number of e-health businesses/companies In the UK especially Ireland we each group or individual simply read out any new details that others may have not seen or researched before.

Practical

Week 3: in the week three practical we were give a series of matlab questions around about 30 of them which was pleasant. These questions basically asked us to input a number of statements to get a response. I took it upon myself to copy and paste each statement into matlab but realised this didn’t work to well as matlab doesn’t like it when you past some statements it will feed back to the user error messages for certain things like Q5 which was e = d’ it doesn’t like the apostrophe and will feed back Error: The input character is not valid in MATLAB statements or expressions.  The first few questions were very straightforward, asking us only to enter in statements to get a reply such as:

4.)

>> d = [

1 2 3

4 5 6

7 8 9]

d =

1     2     3

4     5     6

7     8     9

After creating a number of matrix we where presented with a few questions on matrix manipulation to try and see what happens to different calculations when added together or changed e.g. We can transpose statements such as >> a = a’.

The forth exercise was to try a number of command which I tried and most managed to work e.g

>>mean_a = mean(a);

>> mean_b = mean(b);

>> mean_c = mean(c);

>> mean_all = [mean_a mean_b mean_c]

mean_all =

3     4     5

>>

The next exercise was to visulise data buy crating maps and labled variables using ‘plot’ commands on the figures we downloaded from webct.

The end result from finger 1,2 and 3:

Week 4: Audio signal processing: Is the intentional alteration of auditory signals, or sound. As audio signals may be electronically represented in either digital or analogue format, signal processing may occur in either domain. Analog processors operate directly on the electrical signal, while digital processors operate mathematically on the binary representation of that signal.

Using gold wave we created a speech sample for ourselves in cd quality using a microphone/headset. After this we loaded a few different task which were sound that we must extract breathing sounds from by elimination background sound and noise which we had to do on out own for the final task. Task one and two was just to record our own speech and resample is a different streams.

We also looked into a case study and number of origins to the electronic stethoscope. The ES electronically amplifies body sounds. They require conversion of acoustic waves to electrical signals to be optimized for optimal hearing.

Reference: http://en.wikipedia.org/wiki/Electronic_prescribing

http://en.wikipedia.org/wiki/Audio_signal_processing#Application_areas

Lectures on Monday mornings aren’t always easy but I found this one to be ok. We went through the details on death prevention and leading studies into the causes and statistics of death and how technology has helped in stabling patients and preventing unnecessary pain or death within certain areas. The lecture entailed allot of statics analysis into the leading causes of death and how technology can be used to avoid or reduce suffering etc with patients. We went through the changing demographics and population pyramids to see in what form technology has allowed our population to grow larger and older as well as studying how advances have help clinicians and doctors in patient diagnostics communications and different levels of health technologies such as:

• Tele-medicine
• Tele-health
• E-health
• M-health
• Tele-care
• E-visits

I also gained allot of information about healthcare in providing advances in procedures for protecting patient data and information as well as procedures into medical and surgical practises and finally rehabilitating and easier disease management tactics.

Tutorial:

MatLab means Matrix Laboratory.

We carried on through a number of Vectors and Matrices practices to gather a better understanding of how mat lab works. It made its way to be confusing but in the end I seemed to get how it worked as we were practicing with columns and rows and how using certain commands will impact the output. In some ways it reminded me of the Linux fedora system commands we used last year for systems software.

Played about with a few matlab feature and number columns such as:

>> a = [ 1 2; 2 1 ]

a =
     1     2
     2     1

>> a*a

ans =

     5     4
     4     5

Practical:

Telecare.org – supports a variety of pages to inform patients or relatives/friends of patients on care aid they are able to take to make their lives easier while still maintaining their own independence.

Telehealth refers only to the provision of clinical services whereas telehealth can refer to both clinical and non-clinical aspects.

Clinical uses – Transmission of medical data, health advice over telephone and video conferencing.

Nonclinical – remote admission, internet based healthcare systems, online info, admin uses and healthcare system integration.

Telecare.org has a page detailing equipment to explain to patients/users how the use of sensors and alarms as well as phones and in home systems like pill dispensers can help patients manage their healthcare independently.

Telecare.org quote – More than 90% of people say they want to live independently in their own homes for as long as they possibly can.

http://www.telehealthsolutions.co.uk/

founded in 2006 to converge technology with clinical practice telehealth solutions is a company that has build systems such as the surgery pod and the home pod.

Surgery pod – Allows for announcements to patients with clinicians are available, provide feedback for patients and clinicians as well as the ability to streamline running of specific surgery days and general surgery practises.

Home pod- Home based system for patients to communicated via 3G and provide real independence for patients with technologies such as touch screen systems to dispense pill and proper dosages as well as alert patient to GP appointments etc.

Local Tele-care Company =

Aid Call Ltd Trading as Age Concern Aid Call

Customer contact details

Jenny Pickvance
tel: 01364 651436
email: jenny.pickvance@aidcall-alarms.co.uk
website: www.aidcall.co.uk (opens in a new window)

right well as a part of my health informatics course i am asked to write one of these blog things that people seem to be somewhat interested in these days. first of all i shall explain the first week of semester one in four stages:

Lecture: Managed to make every lecture in first week which is a great success as the university is monitoring attendance like crazy, which is strange as I’m only paying them three and a half grand a year so i should be able to come in when i think its necessary. Anyway we learnt  allot about the health care industry working with technology that we may develop in the future and we also found that health informatics the the first module as a part of two the other being emerging healthcare technologies which i have also picked to do next semester….. oh joy!

Tutorial: Im pretty sure i did the tutorial haha, but it, as most tutorials are its pretty much a lecture in a smaller room. I’m pretty sure that this is the module that ill like the best this semester although some programming aspects are present in the form of matlab but it seems dead-on and not hugely difficult in some form although it has the ability to get difficult.

Practical: wasn’t on…… nice one. But guess who showed up anyway.

Assignment Progress: I like the idea of these blogs i did then in semester one of year two for e-commerce and i intend to use word press to do each blog. we received some details on the proper assignment have been made available to us.

Thanks!

Knowledge management has been implemented by ebuyer in their collaboration with ‘Order-Work’. Although ebuyer confirmed job cuts after this partnership, ebuyer has improved the management of products and services with ‘order-work’ by there side to install products on their online store. At the moment this only applies to business-to-business products but ebuyer is applying their tactics to have this apply also to its normal customer range before Christmas.

‘Dave Wood’ (Ebuyer’s head of commercial operations) – “As digital technology becomes ubiquitous, so has the frustration surrounding installing hardware, particularly routers.
“We wanted to ease that by offering reasonably priced services, but realized it was not viable to provide installation and support services ourselves. ‘Order-Work’s’ business model and values fitted that of Ebuyer.”

Reference – www.channelweb.co.uk

The first of Ebuyer’s products to be promoted were wireless routers, which minimizes customer frustrations when getting connected. I believe this was a great example of knowledge management as is show a long over due insight and improvement to Ebuyer’s on going problem of customer service.

‘Toby Strauss’ (Executive chairman of Order-Work) – “We are delighted to be working with ebuyer to help them expand into services consistent with their reputation for low prices, swift delivery and excellent customer support. ‘Order-Work’s’ flexible resourcing model provides nationwide coverage for ebuyer customers for installs and upgrades of their hardware and software purchases so they can get the most out of their technology.”

Reference – www.allbusiness.com

Ebuyer have made and promoted an increasing knowledge into the way there content has been developed and added positive responses t working with ‘order-work’ especially when dealing with other business and selling wireless products as well as developing and managing the organization better to facilitate large amounts of customers at a satisfactory level. I think its fair to say that the employees have benefited form the collaboration but the few numbers that suffered from job loss have not. Ebuyer’s business environment will allow its employees to change and adapt to company changes and grow with new and relevant ideas to solve different problems.

The main aim that most sites start with is that they want to be ‘famous, popular, well-known and respected’. To gain these attributes they need followers and an audience that will know and understand the website and pass the message of that website onto another selection of people to gain a wide interest in one website. I think it would be fair to say that Ebuyer has gain this following, although not always receiving the most positive reviews, it still is well-know and trusted by its customers and business partners. Although gain reviews, usually stating ‘great prices, rubbish customer service’ it is one of the most well known electronics websites on the internet (for the UK) and is no stranger to visitors. Through searching familiar websites and promotion sites I haven’t been able to find any external marketing that Ebuyer have used to promote themselves. I believe Ebuyer have got to the point whereby their attempts to promote themselves further is futile. They have gain all the popularity they have needed and pulled their customers in. to do this I think it would be obvious to say that Ebuyer started by providing genuine value and letting the ‘word of mouth’ system do the rest, allowing Ebuyer to spread among customers through recommendations.

Objectives:

· Create valuable content. – does your content make sense and it is properly formatted to be read by millions of people? What effect does the business want to have on the customer there can’t just be a constant churning of information. Everything must be frequent and well balanced.

· Create original content. – Everything the business displays on their website must be original and their own content. If a site is built using others content they tend to do well at the start of their life and then go down-hill, meaning original content build for long-term foundations and makes for a more solid and traffic driven website.

· Create timeless content. – allot of the content that websites provide is limited, obscure and out-of-date, this should be avoided by making sure that the audience reading the information are kept well with relevant and ensured information.

· Write for human beings first, computers second. – I can’t say Ebuyer does this objective at all, due to the large amount of customer service complaints it receives and being a company dealing in computers. Therefore this objective is useless.

Visitors:

As a visitor, a reviewer and an analyst of Ebuyer it is hard to determine and even find out the general hits-per-day of the website. But it is common knowledge that Ebuyer and allot of transaction, e-commerce websites like it all gain an excellent source of web traffic from the fact that they have a search engine. If a user stumbles upon the website they are prompted with questions that they can ask the website themselves. Advertising is one easiest ways to get targeted traffic to your website as visitors can visit even by accidently clicking an advertisement on another website or promotion page. Some may even just bookmark the businesses website, which will allow for targeted trafficking.

Ebuyer

Results: 1 – 10

Search Engine – Google

Ebuyer On Google

· www.ebuyer.com

· www.ebuyer.com/googlecheckout

· www.kelkoo.co.uk/mp-3436823-ebuyer

· www.kelkoo.co.uk/m-3436823-ebuyer.html

· en.wikipedia.org/wiki/Ebuyer

· forums.ebuyer.com/

· www.ciao.co.uk/Ebuyer__5107918

· www.pcpro.co.uk/phpbb/viewforum.php?f=29

· www.channelregister.co.uk/2007/07/07/ebuyer_runs_site_on_commodore64/

· static.ebuyer.com/customer/help/index.html?action=YWJvdXRfdXM=

The first result is Ebuyer.com and it displays allot of sub links to pages of the website. The second is a team up with Google allowing Ebuyer to sell its products using goggles new checkout system. The third and fourth websites, a full and sub-website, review other websites and display their best functions.

· Fifth Site – Wikipedia: Provides all info about Ebuyer

· Sixth Site – Ebuyer Forum

· Seventh Site – Ciao: Website review Site

· Eighth Site – PC Pro: External Forum Site

· Ninth Site – Channel Register: Statistics and Development Testing Site

· Tenth Site – Ebuyer terms of service and customer information

One website that did interest me when I got around page four was another service that Ebuyer offered, which was Ebuyer-Software – http://www.ebuyersoftware.com/. Until recently I didn’t know Ebuyer sold or had created their own software programs, for example Anti-Virus and Internet Security as well as an office suite and a photo print package.

Ebuyer On Google

Competitors:

Search Engine – YAHOO

http://www.onino.co.uk/monitors-shrink/37185.html - this seemed to be an outsourced website that bought then sold Ebuyer’s products from their own website. This seemed strange as the website seems to sell its own products and there is only one sections that sells Ebuyer televisions, whether that’s legal or not?!

· When searching on AltaVista the website – http://www.pricerunner.co.uk/sp/ebuyer_keyboard.html

· Compares Ebuyer’s prices to other websites and displays Ebuyer products and prices.

Finally I found another website selling Ebuyer’s products – http://www.nextag.co.uk/ebuyer-a/zzukzB1z0–search-html?nxtg=65e40a280514-E9B62F40490D4F3A. This website sells some Ebuyer laptops. This was another of five all together WebPages that are selling Ebuyer’s products from their own WebPages.

The buzzword: Describes the direction in which information infrastructure seems to be moving. Vast computing resources will reside somewhere and we can connect to them and use them as needed.  Rather than draw from your desktop computer your IT power is delivered of the internet and the general concept incorporates software as a service to satisfy the computing needs of the user.

·         ‘Cloud’ is a metaphor for the internet when depicted in computer network diagrams.

·         Style of computing I which ICT related potential is promoted as a service and it allows users to access technology services from the internet without a expansive knowledge.

IEEE Internet Computing states – “Cloud Computing is a paradigm in which information is permanently stored in servers on the Internet and cached temporarily on clients that include desktops, entertainment centres, table computers, notebooks, wall computers, handhelds, sensors, monitors, etc.”

Bobby Johnston stated in the guardian newspaper that:  “The interesting thing about cloud computing is that we’ve redefined cloud computing to include everything that we already do,” he said. “The computer industry is the only industry that is more fashion-driven than women’s fashion. Maybe I’m an idiot, but I have no idea what anyone is talking about. What is it? It’s complete gibberish. It’s insane. When is this idiocy going to stop?”

Problems:

P: Growing concern that mainstream cloud computing adoption could present a mixture of privacy and ownership issues but at the same time locking consumers out of their own files.

P: Losing control when using proprietary programs. Users need to make sure to do their own computing on their own computer as if users use a proprietary program on somebody else’s computer it leaves them defenceless.

Impact on E-Commerce:

To understand cloud with ecommerce we need to look at transaction websites or secure WebPages. The main one would be banking websites, but in most cases banking WebPages allow users only to access from one computer generally. ‘Google is the owner of a massive computer infrastructure that people tap into from their homes or offices. Recently Google even joined with IBM to promote cloud computing – and for anyone familiar with the history of information technology, that should be a bright red flag’.

Reference:  www.technology.timesonline.co.uk