{"id":464493707291,"title":"(Video) Sequential Data Analysis Assessing Performance of Species Classification (Joe Szewczak)","handle":"video-sequential-data-analysis-assessing-performance-of-species-classification-joe-szewczak","description":"\u003cdiv class=\"page\" title=\"Page 31\"\u003e\n\u003cdiv class=\"section\"\u003e\n\u003cdiv class=\"layoutArea\"\u003e\n\u003cdiv class=\"column\"\u003e\n\u003cstrong\u003eUsing Sequential Data Analysis to Assess the Performance and Limitations of Acoustic Species Classifications\u003c\/strong\u003e\u003cbr\u003e\u003cem\u003eSzewczak, Joseph M.\u003c\/em\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\n\u003cspan\u003e\u003cbr\u003e\u003cbr\u003e \u003c\/span\u003e\u003cspan\u003eHumboldt State University, Arcata, USA\u003c\/span\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003eRuntime: 21:45\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003eRecorded at the 2nd International Symposium, Tucson, AZ 2017\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003eAvailable as a 1280x720 streaming video on demand\u003cbr\u003e\u003cbr\u003e\u003cbr\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\u003cspan\u003e\u003c\/span\u003e\u003c\/div\u003e\n\u003ch4 class=\"column\"\u003e\u003cspan\u003eSummary:\u003c\/span\u003e\u003c\/h4\u003e\n\u003cdiv class=\"column\"\u003e\n\u003cspan\u003eEffective acoustic classification of bat echolocation calls requires a robust and thoroughly inclusive library of species-known recordings to train a classification system to recognize unknown recordings from the field. Species with acoustically similar calls require more samples to establish confident classifications compared with dissimilar pairings, but how many? Sequential data analysis can address this uncertainty and provide an indication of the completeness or incompleteness of library sample size. For example, the acoustically similar congeners \u003c\/span\u003e\u003cspan\u003eMyotis sodalis \u003c\/span\u003e\u003cspan\u003eand \u003c\/span\u003e\u003cspan\u003eM. lucifugus present \u003c\/span\u003e\u003cspan\u003econfoundingly similar calls. A set of 10,955 species-known call samples recorded from tracked individuals in twelve states across the range of \u003c\/span\u003e\u003cspan\u003eM. sodalis \u003c\/span\u003e\u003cspan\u003eenabled a test of the effect of sample size upon classification performance and adequacy of coverage. Classifiers built from randomly selected subsets from 614 to 10,005 call samples were evaluated for classification performance using both the full SonoBat version 4 time-frequency and time-amplitude parameters and AnaLook-equivalent time-frequency parameters. Classification performance ranged from 95.0–69.9% correct using the full SonoBat parameter sets and \u003c\/span\u003efrom 88.6–56.9% correct for the AnaLook-equivalent parameter sets. Both approaches revealed a greater range of performance for smaller data sets, a downward trend in classification performance with larger data sets, and both extrapolated to a meaningless 50% correct performance near 21,000 calls. The range in classification performance with sample size indicated the stochasticity and potential for false certainty inherent in constructing classifiers from inadequate sample sizes. However, this approach also revealed a trend in convergence with increasingly larger subsets that provides a supportive utility for indicating the adequacy of sample size and limitation of classification between ambiguous species. \u003cbr\u003e\u003cbr\u003e\u003cbr\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\u003c\/div\u003e\n\u003ch4 class=\"column\"\u003eThis is an instant download . . .\u003c\/h4\u003e\n\u003cdiv class=\"column\"\u003eImmediately after purchasing this video, you will receive an email with \"Your Instant Download from Bat Conservation and Management \" as the subject, and it will contain download or streaming instructions. If you do not see this email, please check your spam\/junk email folder.\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e","published_at":"2017-12-20T20:04:36-05:00","created_at":"2017-12-20T20:06:05-05:00","vendor":"Bat Conservation and Management, Inc.","type":"Video","tags":["Conference Presentations"],"price":0,"price_min":0,"price_max":0,"available":true,"price_varies":false,"compare_at_price":null,"compare_at_price_min":0,"compare_at_price_max":0,"compare_at_price_varies":false,"variants":[{"id":5389700857883,"title":"Default Title","option1":"Default Title","option2":null,"option3":null,"sku":"","requires_shipping":false,"taxable":true,"featured_image":null,"available":true,"name":"(Video) Sequential Data Analysis Assessing Performance of Species Classification (Joe Szewczak)","public_title":null,"options":["Default Title"],"price":0,"weight":0,"compare_at_price":null,"inventory_quantity":-1,"inventory_management":null,"inventory_policy":"deny","barcode":""}],"images":["\/\/cdn.shopify.com\/s\/files\/1\/2378\/9221\/products\/SequentialDataAnalysis1.jpg?v=1513820179","\/\/cdn.shopify.com\/s\/files\/1\/2378\/9221\/products\/SequentialDataAnalysis2.jpg?v=1513820183","\/\/cdn.shopify.com\/s\/files\/1\/2378\/9221\/products\/SequentialDataAnalysis3.jpg?v=1513820187","\/\/cdn.shopify.com\/s\/files\/1\/2378\/9221\/products\/SequentialDataAnalysis4.jpg?v=1513820192"],"featured_image":"\/\/cdn.shopify.com\/s\/files\/1\/2378\/9221\/products\/SequentialDataAnalysis1.jpg?v=1513820179","options":["Title"],"content":"\u003cdiv class=\"page\" title=\"Page 31\"\u003e\n\u003cdiv class=\"section\"\u003e\n\u003cdiv class=\"layoutArea\"\u003e\n\u003cdiv class=\"column\"\u003e\n\u003cstrong\u003eUsing Sequential Data Analysis to Assess the Performance and Limitations of Acoustic Species Classifications\u003c\/strong\u003e\u003cbr\u003e\u003cem\u003eSzewczak, Joseph M.\u003c\/em\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\n\u003cspan\u003e\u003cbr\u003e\u003cbr\u003e \u003c\/span\u003e\u003cspan\u003eHumboldt State University, Arcata, USA\u003c\/span\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003eRuntime: 21:45\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003eRecorded at the 2nd International Symposium, Tucson, AZ 2017\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003eAvailable as a 1280x720 streaming video on demand\u003cbr\u003e\u003cbr\u003e\u003cbr\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\u003cspan\u003e\u003c\/span\u003e\u003c\/div\u003e\n\u003ch4 class=\"column\"\u003e\u003cspan\u003eSummary:\u003c\/span\u003e\u003c\/h4\u003e\n\u003cdiv class=\"column\"\u003e\n\u003cspan\u003eEffective acoustic classification of bat echolocation calls requires a robust and thoroughly inclusive library of species-known recordings to train a classification system to recognize unknown recordings from the field. Species with acoustically similar calls require more samples to establish confident classifications compared with dissimilar pairings, but how many? Sequential data analysis can address this uncertainty and provide an indication of the completeness or incompleteness of library sample size. For example, the acoustically similar congeners \u003c\/span\u003e\u003cspan\u003eMyotis sodalis \u003c\/span\u003e\u003cspan\u003eand \u003c\/span\u003e\u003cspan\u003eM. lucifugus present \u003c\/span\u003e\u003cspan\u003econfoundingly similar calls. A set of 10,955 species-known call samples recorded from tracked individuals in twelve states across the range of \u003c\/span\u003e\u003cspan\u003eM. sodalis \u003c\/span\u003e\u003cspan\u003eenabled a test of the effect of sample size upon classification performance and adequacy of coverage. Classifiers built from randomly selected subsets from 614 to 10,005 call samples were evaluated for classification performance using both the full SonoBat version 4 time-frequency and time-amplitude parameters and AnaLook-equivalent time-frequency parameters. Classification performance ranged from 95.0–69.9% correct using the full SonoBat parameter sets and \u003c\/span\u003efrom 88.6–56.9% correct for the AnaLook-equivalent parameter sets. Both approaches revealed a greater range of performance for smaller data sets, a downward trend in classification performance with larger data sets, and both extrapolated to a meaningless 50% correct performance near 21,000 calls. The range in classification performance with sample size indicated the stochasticity and potential for false certainty inherent in constructing classifiers from inadequate sample sizes. However, this approach also revealed a trend in convergence with increasingly larger subsets that provides a supportive utility for indicating the adequacy of sample size and limitation of classification between ambiguous species. \u003cbr\u003e\u003cbr\u003e\u003cbr\u003e\n\u003c\/div\u003e\n\u003cdiv class=\"column\"\u003e\u003c\/div\u003e\n\u003ch4 class=\"column\"\u003eThis is an instant download . . .\u003c\/h4\u003e\n\u003cdiv class=\"column\"\u003eImmediately after purchasing this video, you will receive an email with \"Your Instant Download from Bat Conservation and Management \" as the subject, and it will contain download or streaming instructions. If you do not see this email, please check your spam\/junk email folder.\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e"}

(Video) Sequential Data Analysis Assessing Performance of Species Classification (Joe Szewczak)

Product Description
Using Sequential Data Analysis to Assess the Performance and Limitations of Acoustic Species Classifications
Szewczak, Joseph M.


Humboldt State University, Arcata, USA
Runtime: 21:45
Recorded at the 2nd International Symposium, Tucson, AZ 2017
Available as a 1280x720 streaming video on demand


Summary:

Effective acoustic classification of bat echolocation calls requires a robust and thoroughly inclusive library of species-known recordings to train a classification system to recognize unknown recordings from the field. Species with acoustically similar calls require more samples to establish confident classifications compared with dissimilar pairings, but how many? Sequential data analysis can address this uncertainty and provide an indication of the completeness or incompleteness of library sample size. For example, the acoustically similar congeners Myotis sodalis and M. lucifugus present confoundingly similar calls. A set of 10,955 species-known call samples recorded from tracked individuals in twelve states across the range of M. sodalis enabled a test of the effect of sample size upon classification performance and adequacy of coverage. Classifiers built from randomly selected subsets from 614 to 10,005 call samples were evaluated for classification performance using both the full SonoBat version 4 time-frequency and time-amplitude parameters and AnaLook-equivalent time-frequency parameters. Classification performance ranged from 95.0–69.9% correct using the full SonoBat parameter sets and from 88.6–56.9% correct for the AnaLook-equivalent parameter sets. Both approaches revealed a greater range of performance for smaller data sets, a downward trend in classification performance with larger data sets, and both extrapolated to a meaningless 50% correct performance near 21,000 calls. The range in classification performance with sample size indicated the stochasticity and potential for false certainty inherent in constructing classifiers from inadequate sample sizes. However, this approach also revealed a trend in convergence with increasingly larger subsets that provides a supportive utility for indicating the adequacy of sample size and limitation of classification between ambiguous species. 


This is an instant download . . .

Immediately after purchasing this video, you will receive an email with "Your Instant Download from Bat Conservation and Management " as the subject, and it will contain download or streaming instructions. If you do not see this email, please check your spam/junk email folder.
$0.00
Maximum quantity available reached.

"In Stock" items may ship same day to up to 3 business days after placing your order.

"Out of Stock" items may still be purchased, but realize these items may be built-to-order or we may be at the mercy of our vendors for parts of your order or product. Shipping for out of stock items may be days or weeks after placing your order. Some speciaity items may have estimated availability mentioned in their product description above.

All international orders must be paid in full via credit card or international money order before shipping.  We currently accept American Express, Visa, MasterCard and bank/wire transfers.

For international orders, any duties, tariffs or broker fees are not included in the order amount and will be collected by the shipper upon delivery. Any such fees are the responsibility of the customer. 

Please ensure the shipping address you provide can accept UPS deliveries. BCM is not responsible for any fees or delays incurred from an incorrect shipping address.

In Spanish: Todos los pedidos internacionales deben pagarse en su totalidad mediante tarjeta de crédito o giro postal internacional antes de su envío. aceptamos American Express, Visa y Mastercard.  No somos capaces de aceptar banco.

Para los pedidos internacionales, los impuestos, aranceles u honorarios de agente no están incluidos en el orden cantidad y serán recogidos por el cargador a la entrega. estas matrículas son responsabilidad del cliente.        

Por favor garantizar la dirección del envío que usted proporcione puede aceptar envíos ups. bcm no se hace responsable por el pago de honorarios o retrasos de una incorrecta dirección de envío.

What states does sales tax apply?
If you live in PA or FL, you will be charged sales tax according to your state’s sales tax rules.  Tax Exempt Businesses/Organizations: contact us to place your order (717-241-2228). Please have your tax exempt number.

What can I return?
You may return new, unopened items sold within 30 days of delivery for a full refund.

When will I get my refund?
Most refunds are fully refunded in 3-5 business days after we receive and process your return.

Does BCM do replacements?
We can only replace the exact item and version you ordered, and you are responsible to return the damaged item within 30 days. If you want to exchange for a different item, you will need to return the item and place a new order.

Pettersson D500x and M500 Bat Detectors Return and Exchange Policy
All detectors are tested by BCM staff before shipping to customers. Please contact BCM first detailing the issue you are having with the hardware and obtain a Return Authorization. Should email or phone troubleshooting fail to resolve the issue, you should send the unit in suitable packaging to BCM, with the RA number written on the packaging.

BCM will confirm the unit is defective was not abused or water damaged. BCM will replace the unit at our discretion usually within one week. Units requiring repairs may take 6-8 weeks.

Pettersson factory warranty is one year. We have such confidence in this time-tested hardware, BCM extends this warranty 4 additional years for D500x and M500 devices purchased thru BCM, for a total of a 5 year warranty from date of purchase.

D500x External Microphone Return Policy
All microphones are tested by BCM staff before shipping to customers. Please contact BCM first detailing the issue you are having with the Microphone and obtain a Return Authorization. Should email or phone troubleshooting fail to resolve the issue, you should send the unit in suitable packaging to BCM, with the RA number written on the packaging.

DOA microphones are replaced immediately at no charge. There is a $125 for refurbishing used microphones, shipped immediately if in stock.

Electronics
Electronics purchased from BCM are easily returned within 30 days of receipt of shipment. These items must be in new condition with original packaging and accessories. All bat detectors sold by BCM are supported by the device's manufacturers. In the unlikely event a detector appears to be defective, please contact the customer support network for the computer manufacturer to have it serviced under warranty. Pettersson D500x/M500 devices have specific policies, see above.

Software
Software titles purchased from BCM are easily returned within 30 days of receipt of shipment. These items must be unopened and still in their plastic wrap.

General items

Other items purchased from BCM can only be returned if:

  • The return is packaged and shipped in the original packaging.
  • The items are still in new condition.
  • You ship your return within 30 days of receipt of the original shipment.

Any returns that do not meet the above qualifications may receive a reduced refund or be returned to the purchaser. Some items may incur a restocking fee as these items must be inspected, tested, and repackaged. If a restocking fee applies it will be noted on the product information page.

Professional surveys • Artificial Roosts • Equipment

For questions regarding...
New and old orders: email Jenni
Tech questions about cameras, IR lights, bat detectors or SonoBat email John
Bat detectors, repairs, and SonoBat, email Bryan
Professional surveys, quotes, and study plans, email Todd
Bat houses, installations, and harp traps, email Daniel
Shop address: 1263 Claremont Drive • Carlisle, Pennsylvania 17015

Phone: (717) 241-ABAT (2228)
Hours: 9:30 AM-4:00 PM Eastern 

Related Products