Features of MaxDEA Ultra

²  The surprising and exciting thing about MaxDEA Ultra, is that MaxDEA Ultra runs faster - much faster, due to optimized design and parallel computing.

²  MaxDEA Ultra can make full use of your multi-core CPU.

·           With a dual-core CPU, the Bootstrap model may have 50 times speed, the Game Cross Efficiency model may have 8 times speed, other models may have 2 times speed; and

·          With a quad-core CPU, the Bootstrap model may have 100 times speed, the Game Cross Efficiency model may have 16 times speed, other models may have 4 times speed.

²  All functions and feature of MaxDEA Ultra:

·          Easy to use. It needn’t installation and has user-friendly interface. It is very easy to prepare the dataset. You needn’t indicate what are the inputs and outputs by field (variable) names or special arrangement of your data.

·          Easy to backup your DEA models and dataset. Everything is saved in a single file. The software, your dataset and the settings for your DEA model are all integrated into a single Access database file (.mdb), and it is the only file needed for MaxDEA Ultra, so it is very convenient to backup. After closing and reopening MaxDEA Ultra, your database and model settings are still there unchanged.

·          No limitation on the number of DMUs and most comprehensive DEA models.

·          Multiple models can be run at the same time. You can rename or copy the MaxDEA Ultra file freely. Each copy of the file contains one DEA model with all your data and settings saved in the file. You can open and run multiple files simultaneously, taking full advantage of your multi-core CPU. It’s very useful for time-consuming analysis such as bootstrapping.

·          Most important of all, MaxDEA Ultra provides nearly all the possible combinations of up-to-date DEA models. To use a combination of multiple DEA models, just choose all the relevant options. For example, Network-Malmquist model with weakly disposable bad outputs can be achieved by choosing the settings for Network, Undesirable outputs, Weak disposability and Malmquist, at the same time.


DEA Models available in MaxDEA Ultra

MaxDEA Ultra has the most comprehensive DEA models and nearly all their possible combinations, such as the combination of “Undesirable Outputs” and “Malmquist” (Malmquist-Luenberger Productivity Index).

l  Distance to measure efficiency

1)        Radial (CCR 1978; BCC 1984)

2)        Maximum Distance to Frontier (ERM, Enhanced Russel Measure, Pastor, Ruiz, and Sirvent 1999;  SBM, Slacks-based Measure, Tone 2001)

3)        Minimum Distance to Weak  Efficient Frontier (Charnes, Roussea, and Semple1996)

4)        Minimum Distance to Strong Efficient Frontier (Aparicio, Ruiz, and Sirvent 2007)

5)        Directional Distance Function (Chambers, Chung, and  Färe 1996; Chung, Färe, and Grosskopf 1997)

Direction Vector can be

a) Value of the evaluted DMU (x0, y0)

b) Mean of All DMUs

c) Vector  (1, 1,  ……, 1)

d) Range (RDM, Portela, Thanassoulis, and Simpson 2004)

e) Customized (same for all DMUs)

f) Customized (DMU specific)

g) Direction Vector Scanning

6)        A Series of Weighted Additive Models

a) Simple Additive model Weights = (1, 1, 1, ...)

b) Normalized Weighted Additive (Lovell and Pastor 1995)

c) Weights = 1/x0, 1/y0

d) Weights = 1/(mean of x0),  1/(mean of y0)

e) Range Adjusted Measure (RAM, Cooper, Park, and Pastor 1999)

f) Bounded Adjusted Measure (BAM, Cooper, Pastor, Borras, Aparicio, and Pastor 2011)

g) Directional Slacks-based Measure (DSBM, Fukuyama and Weber 2009)

h) Customized Weights (same for all DMUs)

i) Customized Weights (DMU specific)

 

7)        Hybrid Distance(Radial and SBM Fields)

8)        Hybrid Distance(Radial and SBM Measure): (EBM, Epsilon-based Meaure,Tone and Tsutsui 2010)

9)        Cost /Revenue / Profit / Revenue-cost ratio

l  Orientation to measure efficiency

ü  Input-oriented

ü  Output-oriented

ü  Non-oriented

ü  Input-oriented (modified)

ü  Output-oriented (modified)

ü  Non-oriented (input-prioritized)

ü  Non-oriented (output-prioritized)

ü  Non-oriented (generalized priority)

l  RTS to measure efficiency

1)        Constance returns to scale (CRS)

2)        Variable returns to scale (VRS)

3)        Non-increasing returns to scale (NIRS)

4)        Non-decreasing returns to scale (NDRS)

5)        Generalized returns to scale (GRS)

l  Ftontier type to measure efficiency

1)        Convex: Data Envelopment Analysis, DEA (RTS available: CRS, VRS, NIRS, NDRS, GRS)

2)        Non-convex: Free Disposal Hull, FDH (RTS available: CRS, VRS, NIRS, NDRS, GRS)

3)        Non-convex: Elementary Replicability Hull, ERH (Agrell and Tind 2001)

4)        Non-convex: Free Replicability Hull, FRH (Tulkens 1993; Agrell and Tind 2001)

l  Productivity measure (Malmquist index)

a)        Adjacent Malmquist

b)        Fixed Malmquist

c)        Global Malmquist

d)        Sequential Malmquist

e)        Window-Malmquist (Adjacent)

f)          Window-Malmquist (Fixed)

Malmquist index decomposition:

Efficiency Change (catch-up), Technological Change (frontier shift), Scale Efficiency Change, biased Technological Change, TC=OBTC*IBTC*MATC (Fare et al 1997)

l  Window model

 

l  Cluster model

a)        Self-benchmarking

b)        Cross-benchmarking

c)        Downward-benchmarking

d)        Upward-benchmarking

e)        Lower-adjacent-benchmarking

f)            Upper-adjacent-benchmarking

g)        Window-benchmarking

 

l  Customized reference ret model

1)        Variable-benchmark model

2)        Fixed-benchmark model

 

l  Other models

1)        Common Weights Model (Pareto optimal satisfaction degree by Wu, Chu, Zhu, Li, and Liang 2016)

2)        Minimum Efficiency model (Pessimistic DEA) and Interval DEA (Entani, Maeda, and Tanaka 2002)

3)        Network model (Based on the framework by Tone, and Tsutsui 2009; Also Ref. to Tavana, Mirzagoltabar, Mirhedayatian, Saen, and Azadi 2013)

4)        Two-Stage Network (Inputs --> Intermediates --> Outputs) (Kao and Hwang 2008)

5)        Parallel Network (Kao 2009)

6)        Dynamic model

7)        Context-dependent (Seiford, and Zhu 2003)

8)        Super-efficiency model

9)        Modified SBM (Sharp et al 2007)

10)    Cross efficiency model

Second-stage methods are available:

Maximize/Minimize the trade balance of other DMUs as a whole

a) Blanket Benevolent (Type I in Doyle and Green 1995)

b) Blanket Aggressive (Type I in Doyle and Green 1995)

Maximize/Minimize the cross-efficiency of other DMUs as a whole

c) Blanket Benevolent (Type II in Doyle and Green 1995)

d) Blanket Aggressive (Type II in Doyle and Green 1995)

Maximize/Minimize the cross-efficiency of each of other DMUs one by one

e) Targeted Benevolent (Type IV in Doyle and Green 1995)

f) Targeted Aggressive (Type IV in Doyle and Green 1995)

11)    Game Cross Efficiency model  (Liang, Wu, Cook, & Zhu, 2008; Wu, Liang, & Chen, 2009)

12)    Undesirable output model

13)    Nondiscretionary input/output model (non-controllable model, measure specific model)

14)    Bounded input/output model

15)    Preference (weighted) model

16)    Restricted projection model

17)    Weak disposability model

18)    Restricted multiplier model (assurance region model, trade-offs between inputs and outputs)

19)    Fuzzy DEA

20)    MetaFrontier DEA (Rao, O'Donnell, and Battese 2003)

21)    Non-concave MetaFrontier DEA (Tiedemann, Francksen, and Latacz-Lohmann 2011)

22)    Non-concave MetaFrontier DEA and Non-concave MetaFrontier Malmquist

23)    Bootstrap

a)        Bootstrap of DEA Score

b)        Bootstrap of Malmquist Index